Modelos termo-mecánicos para las zonas de subducción de

Transcription

Modelos termo-mecánicos para las zonas de subducción de
UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO
PROGRAMA DE POSGRADO EN
CIENCIAS DE LA TIERRA
INSTITUTO DE GEOFISICA
MODELOS TERMOMECÁNICOS PARA LAS ZONAS
DE SUBDUCCIÓN DE GUERRERO Y KAMCHATKA
TESIS
Para obtener el grado de
DOCTOR EN GEOFÍSICA
(SISMOLOGÍA Y FÍSICA DEL INTERIOR DE LA TIERRA)
PRESENTA
Vlad Constantin Manea
2004
A Dr. Vladimir Kostoglodov,
por todo el apoyo que me ha brindado desde mi llegada al
Departamento de Sismología
y
por haber puesto su confianza en mí.
Gracias por su amistad.
Agradecimientos
Agradezco a los coordinadores del programa de Posgrado, Drs. Blanca
Mendoza Ortega y Oscar Campos-Enríquez, por todo el apoyo que me otorgaron.
Agradezco a mi comité de sinodales quienes lograron enriquecer este
trabajo con sus comentarios muy atinados, a los Drs. Singh Krishna Singh, Rosa
María Prol Ledesma, Juan Manuel Espíndola Castro, Iouri Taran Sobol, Luca
Ferrari Pedraglio y Luis Delgado Argote.
Gracias a la beca que me otorgo la DGEP y al apoyo que me otorgaron
PAEP, CONACYT (G25842-T, 37293-T), y PAPIIT (IN104801) pude finalmente
concluir el doctorado con este trabajo de investigación.
Mucho aprecio al Dr. Granville Sewell por toda su ayuda en la aplicación de
los métodos numéricos y por su amistad.
Muchas gracias a Dr. Carlos Mortera Gutiérrez por todo el apoyo que me
brindó y por su amistad.
Muchas gracias a Drs. Carlos Valdez y Raul Valenzuela Wang por todo el
apoyo y los buenos consejos que me dieron.
Muchas gracias a Dr. Arturo Iglesias que estuvo siempre cerca como buen
compañero y amigo. Muchas gracias a ing. Manuel Velásquez por todo su apoyo
en todos los problemas técnicos que ocurrieron y por su amistad.
Agradezco a Araceli y a Mónica del Posgrado en Ciencias de la Tierra y a
Paty y Adriana del Departamento de Sismologia, por toda su ayuda con los
tramites burocráticos.
Asimismo, quiero agradecer a Drs. Gillian Foulger, Taras Gerya, Bradely
Hacker, Peter van Keken, Jeffery Park, Dean Presnall, Harro Schmeling, Jeroen
van Hunen y Steven Ward quienes apoyaron en la revisión de los tres articulos
que forman parte de esta tesis.
Sinceramente me siento afortunado por haber tenido la oportunidad de
estudiar en el Posgrado en Ciencias de la Tierra de la UNAM y por haber
convivido con excelentes profesores y compañeros.
INDICE
Resumen .........................................................................................................
1
I. Introducción ..................................................................................................
6
II. Estructura térmica, acoplamiento y metamorfismo en la zona de
subducción mexicana debajo del estado de Guerrero
Thermal Structure, Coupling and Metamorphism in the Mexican Subduction
Zone beneath Guerrero ……………………………………………………………
19
III. Modelo termomecánico de la cuña del manto en la zona de subducción
del centro de México y el mecanismo “blob tracing” para el transporte de
magma
Thermo-mechanical model of the mantle wedge in Central Mexican
subduction zone and a blob tracing approach for the magma transport …….
53
IV. Modelos térmicos, transporte del magma y estimación de la anomalía de
velocidad debajo de Kamchatka meridional
Thermal Models, Magma Transport and Velocity Anomaly Estimation
beneath Southern Kamchatka ......................................................................... 108
V. La sísmicidad intraplaca y los esfuerzos térmicos en la Placa de Cocos
debajo de la parte central de México
Intraslab Seismicity and Thermal Stress in the Subducted Cocos Plate
beneath Central México ................................................................................... 165
VI. Discusión y Conclusiones .......................................................................
193
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
RESUMÉN
La temperatura es una de las propiedades físicas de mayor importancia en
la zona de subducción. La distribución de la temperatura con la profundidad y la
distancia desde la trinchera depende principalmente de los siguientes parámetros:
la forma de la placa subducida y la edad de la misma, la velocidad de
convergencia y los parámetros térmicos de las rocas. Dentro de estos parámetros,
la geometría de la placa subducida es uno de gran importancia. En el mundo
existen básicamente dos tipos de geometrías para la placa subducida: una que
contiene un segmento subhorizontal antes de entrar en la astenósfera (México,
Chile) y otra que buza en la astenósfera a un ángulo de echado mayor y sin tener
este segmento subhorizontal (México, Cascadia, Nicaragua, Costa Rica,
Kamchatka). En este trabajo se han considerado dos placas oceánicas de
subducción con geometrías, edades y velocidades de convergencia diferentes:
una joven de ca. 14 Ma (placa de Cocos frente al estado de Guerrero) y la otra
vieja de ca. 70 Ma (placa Pacífico subducida debajo de la península de
Kamchatka).
Los modelos de temperatura asociados con la zona de subducción son muy
útiles para estudiar fenómenos asociados a la sismicidad y el volcanismo. La
sismicidad debajo de la costa esta relacionada con la temperatura, debido a que
los terremotos en esta región ocurren en donde se encuentra el rango de
temperatura de 100 °C - 250 °C. También los modelos térmicos son de gran
interés para estudiar los reciéntemente descubiertos terremotos “lentos” en
Guerrero (de hecho, los terremotos “lentos” representan una deformación que dura
varios meses y es similar a la deformación observada durante un sismo normal
que dura segundas) que ocurren en un rango de temperaturas de 250 °C - 450 °C.
Además, en el presente estudio, a partir de la estructura térmica, se presenta un
modelo de los esfuerzos termoelásticos en la placa de Cocos subducida debajo la
placa de Norteamérica que podría explicar los eventos sísmicos con mecanismo
normal intraplaca, que ocurren a una mayor profundidad dentro de la placa
-1-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
oceánica subducida. Usando la relación entre la presión y la temperatura para
basalto (diagrama de fase para basalto) y la temperatura en la superficie de la
placa subducida (geoterma), obtenida a través de los modelos térmicos, se puede
proporcionar información sobre los tipos de materiales (secuencias metamórficas)
involucrados en el contacto entre la placa subducida y la placa continental. Otra
rama de estudio derivada del conocimiento del campo térmico es el volcanismo a
través de los modelos físicos para el transporte de magma en la cuña del manto
(Cap. III).
La estructura térmica de una zona de subducción puede ser obtenida
usándose el modelado numérico. Para el modelado térmico se usó un esquema
numérico 2D que resuelve un sistema de ecuaciones compuesto por las
ecuaciones Stokes y la ecuación del flujo de calor. El principal mineral que entra
en la composición mineralogica de la cuña del manto es el olivino, así que en la
preparación de los modelos termomecánicos se tomo en cuenta la reología del
olivino seco. Finalmente, el sistema de ecuaciones fue resuelto con la ayuda del
método de elementos finitos.
En el caso de la estructura térmica de la zona de subducción de México
central, debajo del estado de Guerrero se determinó la distribución de la
temperatura hasta una profundidad de 250 km para un perfil normal a la Trinchera
Mesoaméricana que pasa por Acapulco y que se extiende ~ 600 km dentro del
continente. Los resultados indican que la zona de fuerte acoplamiento debajo de la
costa, en donde ocurren los terremotos de contacto entre las placas, corresponde
a un intervalo de temperatura de 100 º - 250 °C y se extiende hasta una distancia
de ~ 80 km desde la trinchera. Los modelos de deformación que se ajustan bien a
los datos de GPS, muestran que la zona acoplada debajo de Guerrero se extiende
hasta una distancia, de ~ 200 km a partir de la trinchera. Después de esta región
acoplada, los modelos muestran que la Placa de Cocos subducida y la Placa de
Norteamérica están completamente desacopladas. Según datos experimentales,
la temperatura de 450 °C representa la transición hacia un régimen de
deslizamiento libre (Cap. II).
-2-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Posteriormente, con la ayuda del diagrama de fases para basalto se
identificaron y localizaron las secuencias metamórficas en la corteza oceánica
subducida. Se pudo observar que debido a los cambios metamórficos, una fuerte
deshidratación ocurre a lo largo del contacto entre las dos placas, liberando hasta
~ 4 wt% de H2O. Se observó una buena correlación entre los cambios
metamórficos y los elementos con una cierta cantidad de acoplamiento de los
modelos de deformación. La región en donde los terremotos lentos ocurren
(aproximadamente 80 - 200 km desde la trinchera) corresponde a la facies
metamórfica de esquitos azules. Junto con la liberación de agua por
deshidratación, la facies de esquitos azules parece ser responsables de la
ocurrencia de los eventos sísmicos lentos.
La estructura térmica debajo del volcán Popocatépetl muestra una
temperatura máxima de ~ 1300 °C en la cuña del manto, que es suficiente para
fundir la peridotita hidratada con agua que provendría de las reacciones de
deshidratación en la placa subducida. La geoterma de la superficie de la placa
subducida intercepta la curva del “solidus” para basalto a una profundidad de ~ 70
km y para sedimentos saturados a ~ 50 km. Esto sugiere que hay tres fuentes
para producir material magmático debajo de la Faja neo volcánica mexicana: la
peridotita, la placa basáltica y los sedimentos subducidos.
A continuación se desarrolló un modelo físico para el transporte de magma
hasta la base de la corteza continental. En el modelo se propone que el magma es
producido y se acumula en la proximidad de la superficie de la placa subducida
debajo de la faja volcánica formando burbujas con diversos diámetros y
viscosidades. Los resultados muestran que algunas burbujas nunca suben a la
superficie, principalmente debido al diámetro pequeño (~ 0.6 km). Al aumentar el
diámetro a ~ 1 km las burbujas pueden llegar hasta el Moho a través de una
trayectoria sinuosa por la cuña del manto. Las burbujas llegan en lugares distintos,
dependiendo de su diámetro y viscosidad. Si el diámetro aumenta hasta 10 km,
todas las burbujas llegan al mismo punto justamente arriba del punto de
formación.
-3-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
En el capitulo IV, el mismo esquema numérico fue aplicado para la zona de
subducción muy vieja (> 70 Ma) de Kamchatka. El perfil para el modelado de esta
región está ubicado en el sur de Kamchatka, lejos del contacto entre la trinchera
de Kamchatka y el arco de Aleutianas en donde hay una anomalía térmica muy
fuerte. Los modelos muestran una temperatura > 1300 °C debajo del área
volcánica, suficiente para producir los magmas calcialcalinos comunes en
Kamchatka. Un problema complejo es obtener la distribución de la temperatura en
la cuña del manto. En ese sentido es de gran ayuda la tomografía sísmica. En la
parte sur de Kamchatka se usó una relación entre la atenuación de las ondas
sísmicas y la temperatura de fusión del olivino para obtener un modelo de
anomalías de velocidad debajo de la faja volcánica a partir de un modelo térmico.
La anomalía de velocidad obtenida con este método para el Sur de Kamchatka
muestra una anomalía negativa de 7% debajo del área volcánica. Este resultado
concuerda con la anomalía de velocidad en la tomografía sísmica con atenuación
de 7%. Aunque la forma de la anomalía de velocidad es diferente, una buena
correlación con la magnitud podría darnos una indicación sobre una buena
estimación de la magnitud de la temperatura en la cuña del manto. El modelo de
las burbujas flotantes también se utilizó para el sur de Kamchatka, pero en este
caso se calculó numéricamente la historia térmica de una burbuja de 10 km en
diámetro que sube hasta el Moho en 2.6 Ma. El esquema numérico utiliza el
mecanismo de transferencia del calor por conducción entre la burbuja y el manto.
Los resultados muestran que la burbuja alcanza el Moho sin solidificarse, teniendo
para el 90% de ella una temperatura mayor a 900 °C. Este resultado sugiere la
posibilidad de acumulación de estas burbujas en la base del Moho, probablemente
formando una cámara magmática profunda debajo del área volcánica.
Una característica adicional de la zona de subducción en la parte central de
México, es la presencia exclusiva de terremotos cuyos mecanismos focales son
normales (extensión) dentro de la placa oceánica subducida. Como la mayor parte
de estos eventos normales se ubica dentro de la parte subhorizontal de la placa,
hay muy poca influencia en el campo de esfuerzos debido al doblamiento de la
-4-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
placa. La disponibilidad de un modelo térmico para Guerrero se usó para
determinar los esfuerzos termoelásticos dentro de la litosfera oceánica (Cap V).
El esquema numérico usa las ecuaciones clásicas de la teoría de termoelasticidad en placas delgadas. Los resultados muestran una distribución de los
esfuerzos termoelásticos que consiste en tres capas: una capa central
caracterizada por esfuerzos termoelásticos de tensión máximos de 0.3 kbar y dos
capas externas con esfuerzos termoelásticos de compresión máximos de 0.65
kbar. La capa inferior con esfuerzos de compresión se desvanece cuando en la
base de la litosfera se incluye una ley de decaimiento exponencial de los
esfuerzos debido a temperaturas mayores de 700 °C. También, incluyendo un
momento de torsión para la parte horizontal de la placa subducida, la capa
superior de esfuerzos de compresión se desvanece. Este momento de torsión se
produce porque la placa subducida baja en la astenósfera. De este manera se
queda nada mas una capa de esfuerzos de extensión con una magnitud de 0.5
kbar en donde la mayoría de los sismos de extensión se encuentran. La ubicación
de la isoterma de 700°C concuerda bien con la profundidad máxima de los
terremotos intraplaca y se puede considerar como una temperatura de corte. Para
temperaturas mayores a este valor de corte, los terremotos no pueden existir
debido al comportamiento dúctil de la placa oceánica subducida.
Este estudio demuestra que los modelos térmicos para la zona de
subducción proporcionan informaciones útiles para los estudios geodinámicos
relacionados con los eventos sísmicos lentos, la sismicidad entre las placas y la
sismicidad dentro de la placa subducida, a través de los esfuerzos termoelásticos.
También, los modelos de flujo en la cuña del manto permiten estudiar el
comportamiento de los magmas en su trayecto hacia la base de la corteza
continental.
-5-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
I INTRODUCCIÓN
Una de las características más importantes de la zona de subducción
mexicana es la forma subhorizontal de la placa subducida en la parte central,
debajo del estado de Guerrero (Kostoglodov et al., 1996). Otra particularidad
importante de esta area es la presencia de la brecha sísmica de Guerrero, que se
extiende ~ 120 km desde Acapulco hacia el NE, no ha producido ninguna ruptura
desde 1911, mientras que en las zonas cercanas han ocurrido grandes terremotos
de mecanismo inverso.
Observaciones
recientes
de
GPS
en
Guerrero
muestran
que
el
acoplamiento intraplaca durante el periodo intersísmico es anormalmente largo,
con una extensión desde la trinchera, entre 180 y 220 km hacia el continente
(Kostoglodov et al., 2003). Algunos modelos térmicos propuestos para la zona de
subducción de Guerrero (Currie et al., 2002), no toman en cuenta la zona de
acoplamiento de ~ 200 km debido a la falta de esa información en aquellos
tiempos.
Usando la geometría de la placa subducida de Kostoglodov et al. (1996) y
una área grande parcialmente acoplada de cerca de 220 km desde la trinchera, el
modelo de dislocación de Kostoglodov et al. (2003) se ajusta bien con los datos de
deformación observados. Haciendo la comparación del modelo de la dislocación
que usa la geometría de la placa de Currie et al. (2002) con las deformaciones
observadas en la superficie, no se puede observar una buena correlación debido a
la geometría de la placa subducida y a la extensión limitada de la zona
parcialmente acoplada. Nuevos estudios son necesarios para refinar los modelos
existentes y para explicar la zona de acoplamiento de gran extensión (~ 200 km)
desde la trinchera. También, estudios adicionales son necesarios para verificar
una posible correlación entre los cambios metamórficos a lo largo del contacto
entre las placas de Cocos y Norteamérica y las zonas con diversos grados de
acoplamiento que satisfacen los modelos de deformación de Kostoglodov et al.
(2003).
-6-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Los pocos modelos térmicos que hay para la zona de subducción de
México (Currie et al., 2002; Manea et al., 2004a) no presentan una estructura
térmica detallada y confiable porque no toman en la cuenta la reología de la
astenósfera. Los modelos termomecánicos de la cuña del manto con reología
dependiente de la temperatura y/o la tasa de deformación están desarollandose
(Furukawa, 1993; Conder et al., 2002; Van Keken et al., 2002; Kelemen et al,
2003). La dinámica de la cuña del manto y la estructura térmica de las zonas de
subducción subhorizontales someras han sido investigadas recientemente por van
Hunen et al. (2002).
El modelo térmico previo para la zona de subducción de Guerrero con una
interfaz somera de la placa usando una expresión analítica predefinida para el flujo
en el rincón del manto (Currie et al., 2002), predice una temperatura de ~ 900ºC
para la astenósfera debajo del frente volcánico (por ejemplo, el volcán
Popocatépetl), considerada muy baja para producir fundido. En este modelo, el
componente basáltico de la placa subducida tampoco alcanza la temperatura del
fundido. Como consecuencia, este modelo no puede explicar la fuente del magma
para el Cinturón Volcánico Trans-Mexicano (CVTM) en su parte central.
El CVTM es un arco volcánico neógeno construido sobre el borde sur de la
placa de Norteamérica (Ferrari et al., 1999). Moore et al. (1994) es el primero en
sugerir la presencia de una pluma del manto debajo Guadalajara. Este modelo ha
sido expandido mas tarde por Márquez el al. (1999). Ellos proponen que todo el
CVTM esta relacionado con una pluma del manto que impactó la parte oeste de
México en el Mioceno tardío. En el modelo de Márquez et al.(1999) la pluma corta
primero la placa subducida que, a su turno, puede cortar la cabeza de la pluma.
Los dos modelos se apoyan en los datos de geoquímica y no ajustan con la
geología y tectonica del CVTM. En un comentario por Ferrari y Rosas (1999),
acerca del trabajo de Marquez et al. (1999) se señala que: ni el rifting ni los
magmas tipo OIB (Ocean Island Basalt) no presentan el aumento de la edad
requerido por el modelo de la pluma; también, en el oeste de México, en donde la
pluma podría impactar, no hay ninguna evidencia de levantamiento regional;
-7-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
asimismo, el volumen de los magmas OIB en el CVTM representa nada mas una
fracción del volcanismo relacionado con la subducción, y mucho más pequeña que
los flujos basalticos continentales típicos.
Sheth et al. (2000) y Verma (2002) han hecho recientemente comentarios
sobre el papel de la subducción en la génesis del CVTM y propusieron que el
rifting continental encima del manto heterogéneo podría ser el mecanismo de
formación. El rechazo de la influencia de la subducción sobre la génesis del CVTM
se fundamenta por lo general en las bases petrológicas y de estadística
geoquímica, con la ayuda de una interpretación sencilla acerca de la falta de
sismicidad debajo del arco. En este estudio se aclara que la falta de sismicidad
debajo el CVTM es debido a las temperaturas muy altas (> 700 ºC) en la placa
subducida de Cocos.
Recientemente, Cervantes y Wallace (2003) analizaron los elementos
mayores y de traza en las inclusiones fundidas en algunos conos de ceniza en
Sierra Chichinautzin. Encontraron lavas con alto contenido de agua, señalando
que los fluidos de subducción penetran el manto debajo de CVTM. Gómez-Tuena
et al. (2003) encontraron la presencia de un cinturón WNW-ESE con una firma
adakitica, sugiriendo que parte de la placa subducida debajo del CVTM está
fundida. Esta es otra evidencia sobre la influencia de la subducción debajo el
CVTM.
Un modelo reciente por Ferrari (2004) propone un mecanismo basado en el
desprendimiento de la placa subducida y unas infiltraciones de la astenósfera
enriquecida con el fin de explicar la ocurrencia de los magmas de tipo OIB.
Estudios geológicos recientes (Luhr, 1997; Márquez et al., 1999a) sugieren
que hay un manto complejo debajo del CVTM. Sin embargo, el origen de los
basaltos tipo OIB no es muy claro todavía. La advección en el manto astenosférico
debido al buzamiento de la Placa de Cocos ha sido propuesta como una fuente
posible de los magmas tipo OIB (Wallace y Carmichael, 1999). Hay algunos
estudios mineralógicos y geoquímicos sobre el CVTM (e.g. Márquez y De Ignacio,
2002) que sugieren la existencia de dos tipos primitivos de magmas máficos: uno
-8-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
con un componente “astenosférico” tipo OIB y el otro con un componente
“litosférico”. Aunque la corelación entre el volcanismo y la Placa de Cocos es
conocida (e.g. Pardo y Suárez, 1995; Wallace y Carmichael, 1999) hay hipótesis
diferentes con respecto al origen del volcanismo en esta área como: la extensión
del sistema de fallas del Golfo de California (Ferrari et al., 1994), una zona de
corteza vieja debilitada (Cebull y Schubert, 1987), mecanismos de transpresión de
la corteza (Shubert y Cebull, 1984; Ferrari et al., 1990) y rifting (Luhr, 1997;
Márquez et al., 1999a,b).
También, la fuente de los fluidos que metasomatizan el manto no es clara.
Hay
tres
posibilidades
propuestas
como
explicación
para
el
manto
metasomatizado (Márquez et al., 1999c; Wallace y Carmichael, 1999; Verma,
1999; 2000): a) la cuña del manto afectada por los fluidos generados por la
deshidratación de la Placa de Cocos subducida; b) el manto litosférico viejo,
enriquecido; c) el manto metasomatizado por los volátiles de los plumas del
manto. Es imposible obtener evidencias directas acerca de la cuña del manto
metasomatizada por los fluidos liberados de la placa oceánica subducida, aunque
eventualmente, los xenolitos del manto son raras veces encontrados en las lavas
de los arcos volcánicos. En el CVTM las andesitas hornbléndicas cuaternarias que
irrumpen en la proximidad del área de El Peñón (ver la Fig.1, pp.87) contienen
xenolitos de 1 - 2 cm, ricos en fenocristales de hornblenda y con matriz andesítica
empobrecida en plagioclasa (Blatter y Carmichael, 1998). Todas estas
observaciones sugieren un flujo de volátiles, que provienen de la placa subducida,
en la cuña del manto.
Debajo del CVTM la subida del magma hacia la superficie produce grandes
estratovolcanes y volcanes monogéneticos más pequeños esparcidos en áreas
grandes. En su parte central, el CVTM contiene algunos estratovolcanes como
Popocatépetl, Iztaccíhuatl y Nevado de Toluca. Los volcanes monogéneticos están
representados en general por conos de lava, conos de ceniza, así como domos y
flujos de lava. Mientras los estratovolcanes se caracterizan por la periodicidad de
sus erupciones, los volcanes monogéneticos presentan un solo evento eruptivo y
-9-
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
por esto, tienen menor tamaño. Takada (1989, 1994) proponen un mecanismo
doble que relaciona el aporte de magma y el esfuerzo regional.
El relleno sedimentario en la trinchera Mesoamericana es muy pequeño,
sugiriendo que ~ 95% de esto sedimentos entran en subducción (Manea et al.,
2003). Un flujo de volátiles de la placa oceánica metamorfizada y de los
sedimentos puede iniciar la fusión parcial de la peridotita en la proximidad de la
placa subducida (Tatsumi, 1986; Davies y Stevenson, 1992). Gerya y Yuen (2003)
proponen que una instabilidad Raileigh-Taylor desarrollada en la superficie de la
placa subducida puede generar burbujas flotantes hasta de 10 km de diámetro que
pueden penetrar la cuña del manto. Gerya y Yuen (2003) sugieren que la
composición química de estas burbujas está relacionada con el fundido parcial de
la corteza oceánica basaltica y gabroica, el fundido parcial del manto hidratado y
con los sedimentos oceanicos fundidos.
Además de las características de la zona de subducción de Guerrero
relacionadas con la forma de la placa y la extensión de la zona acoplada, hay otra
característica importante: todos los terremotos intraplaca en la zona de Guerrero
tienen mecanismo normal, ocurriendo a partir de ~ 85 km de la trinchera.
Se ha aceptado que la litósfera oceánica fría que se subduce, entra en el
manto más caliente produciendo esfuerzos termoelásticos importantes que
pueden generar terremotos intraplaca. La distribución de los esfuerzos
termoelásticos ha sido propuesta como una explicación para un doble plano de
falla en varias zonas de subducción como Japón y Kamchatka (Hamaguchi et al.,
1983, Gorbatov et al., 1997). La distribución de los esfuerzos termoelásticos según
Hamaguchi et al. (1983) muestra esfuerzos de compresión muy grandes (hasta 10
kbars) para la parte superior e inferior de la placa subducida, y esfuerzos de
tensión (hasta 7.5 kbars) para la parte central de la placa. Estos valores son con ~
2 órdenes de magnitud mayores que la caída de esfuerzos durante los terremotos
intraplaca (~ 0.1 kbars).
Estos valores muy grandes de los esfuerzos termo-elásticos están
relacionados con un gran contraste de temperatura de hasta ∆T = 1000°C dentro
- 10 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
de la placa subducida. La estructura simplificada de la cuña del manto propuesta
por Hamaguchi et al. (1983) está representada por un gradiente térmico de
10°C/km hasta 100 km de profundidad y 5°C/km hasta 150 km de profundidad. Los
modelos termomecánicos más recientes para la cuña del manto con una reología
dependiendo de temperatura y/o esfuerzo (Furukawa, 1993; Conder et al., 2002;
Van Keken et al., 2002; van Hunen et al., 2002; Kelemen et al, 2003, Manea et al.,
2004b) muestran una distribución más complicada de la temperatura. De este
modo, el contraste de temperatura dentro de la placa subducida está controlado
por la estructura térmica de la cuña del manto.
Nuevos estudios acerca del campo de esfuerzos termoelásticos en la zona
de subducción mexicana son necesarios y esto se puede realizar usando los
modelos térmicos recientes, desarrollados para una viscosidad en la cuña del
manto, dependiendo muy fuerte de la temperatura
Además, es necesario un estudio para investigar la relación entre los
esfuerzos termoelásticos y los terremotos normales intraplaca debajo de la zona
de subducción de Guerrero.
Otra zona de subducción importante, considerada como vieja, es la zona de
subducción de Kamchatka. Hasta la fecha no hay un estudio de la distribución del
campo térmico asociado al fenómeno de subducción en esta región, aunque hay
muchos estudios con respecto a la composición del magma a lo largo de la
península (por ejemplo Graybill et al., 1999).
Para investigar con detalle la distribución del campo térmico debajo del
centro del CVTM y sur de Kamchatka, nuevos modelos termomecánicos son
necesarios, incluyendo la reología del olivino (el mineral más abundante en el
manto superior). Además, para proporcionar más información acerca de la fuente
del magma un modelo de propagación del magma a través de la cuña del manto
permitirá mejorar nuestro conocimiento sobre la relación entre el volcanismo y la
dinámica de la zona de subducción.
- 11 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Las metas principales del presente estudio son:
-Estudiar la relación entre el campo térmico y la ocurrencia de los eventos
sísmicos lentos en México;
-Examinar una posible dependencia entre los cambios metamórficos a lo largo del
contacto entre las placas de Cocos y Norteamérica y los modelos de deformación
existentes;
-Obtener la distribución del campo térmico y convectivo debajo del CVTM y de la
parte sur de Kamchatka;
-Estudiar la relación entre las condiciones de P-T debido a los modelos térmicos y
la composición química de los productos volcánicos en superficie;
-Estudiar uno de los posibles mecanismos de transporte del magma desde la
superficie de las placas subducidas (Placa de Cocos y Placa de Norteamérica)
hasta el Moho.
Adicionalmente, este estudio trata de:
-Obtener una imagen tomográfica a partir de los modelos térmicos debajo del sur
de Kamchatka y compararla con la imagen tomográfica ya existente obtenida a
partir de la atenuación de las ondas sísmicas P;
-Calcular los esfuerzos termoelásticos en la Placa de Cocos debajo del estado de
Guerrero y compararlos con la distribución de los sismos intraplaca en la región.
En el capítulo II se presenta un estudio que determinan las relaciones entre
el modelo térmico y la ocurrencia de los terremotos lentos en México. Además,
con la ayuda del diagrama de fases para basalto se investiga la correlación entre
los cambios metamórficos a lo largo del contacto entre las placas y los resultados
del modelado de deformación que satisfacen los datos observados. El capítulo III
se enfoca en la distribución del campo térmico y de velocidades debajo del CMVB.
Se analiza también la propagación del magma hacia el Moho usando un modelo
- 12 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
físico de burbujas flotantes en un manto convectivo. En el capítulo IV se aplica el
mismo esquema numérico para obtener la distribution de la temperatura para el
sur de Kamchatka. Además, el modelo de burbujas flotantes se desarrolla con
mayor detalle para tener un mejor conocimiento sobre la evolución térmica de las
burbujas a través del manto. En este capítulo, la disponibilidad de una tomografía
sísmica para el sur de Kamchatka hace posible constreñir el campo térmico debajo
del CVTM. En el capítulo V se calculan los esfuerzos termoelásticos inducidos por
la distribución nouniforme de la temperatura en la Placa de Cocos debajo del
estado de Guerrero. Se estudia también la correlación entre la distribución de los
esfuerzos termoelásticos y la sismicidad intraplaca en la región. En el capítulo VI
se presentan la discusión y las conclusiones del trabajo en su conjunto.
- 13 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
REFERENCIAS
Blatter, D.L., and Carmichael, I.S.E., 1998. Hornblende peridotite xenoliths from
central Mexico reveal the highly oxidized nature of subarc upper mantle. Geology,
v. 26, no. 11, pp. 1035-1038.
Carmichael, I.S.E., 2002. The andesite aqueduct: perspectives on the evolution of
intermediate magmatism in west-central (105° 99' W) México. Contribution Mineral
Petrology, 143, 641-663.
Cebull, S.E., and Schubert, D.H., 1987. Mexican Volcanic Belt-an intraplate
transform. Geofisica Internacional, v.26, pp. 1-13.
Cervantes P., and Wallace P., 2003. Role of H2O in subduction zone magmatism:
new insights from melt inclusions in high-Mg basalts form central México. Geology,
31, 235-238.
Conder, J.A., Weins, D.A., and Morris, J., 2002. On the decompression melting
structure at volcanic arcs and back-arc spreading centers. Geophysical Research
Letters, vol. 29, 17-1-17-4.
Currie, C.A., Hyndman, R.D., Wang, K., and Kostoglodov, V., 2002. Thermal
models of the Mexico subduction zone: Implications for the megathrust
seismogenic zone. Journal of Geophysical Research, 107, no. B12, 2370, doi:
10.1029/2001JB000886.
Davies, J.H., and Stevenson, D.J., 1992. Physical model of source region of
subduction zone volcanism. Journal of Geophysical Research, v. 97, p. 2037-2070.
Ferrari, L., Pasquare, G., and Tybaldi, A., 1990. Plio-Quaternary tectonics on the
central Mexican Volcanic Belt and some constraints on its rifting mode. Geofisica
Internacional, v. 29, pp. 5-18.
Ferrari, L., Garduno, V., Innocenti, F., Manetti, P., Pasquere, G., and Vaggeli,
G., 1994. Volcanic evolution of central Mexico: Oligocene to Present. Geofisica
Internacional, 33, 91-105.
Ferrari L., and Rosas-Elguera, J., 1999. Alkalic (OIB type) and calc-alkalic
volcanism in the Mexican Volcanic Belt: a case for plume-related magmatism and
- 14 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
propagating rifting at an active margin? Comment to the article by Marquez, A.,
Oyarzun, R., Doblas, M., Verma, S. P. Geology, 27, 1055-1056.
Ferrari L., Conticelli S., Vaggelli C., Petrone C., and Manetti P., 2000. Late
Miocene mafic volcanism and intra-arc tectonics during the early development of
the Trans-Mexican Volcanic Belt. Tectonophysics, 318, 161-185.
Ferrari L., 2004. Slab detachment control on volcanic pulse and mantle
heterogeneity in Central Mexico. Geology, v. 32, p. 77-80.
Furukawa, F., 1993. Magmatic processes under arcs and formation of the volcanic
front. Journal of Geophysical Research, 98, 8309-8319.
Gerya, T.V. and Yuen, D.A., 2003. Rayleigh-Taylor instabilities from hydration and
melting propel ‘cold plumes’ at subduction zones. Earth and Planetary Science
Letters, v. 212, p. 47-62.
Gómez-Tuena A., LaGatta A., Langmuir C., Goldstein S., Ortega-Gutierrez F.,
and Carrasco-Nuñez G., 2003. Temporal control of subduction magmatism in the
eastern Trans-Mexican Volcanic Belt: mantle sources, slab contributions and
crustal
contamination.
Geochemistry,
Geophysics
and
Geosystems,
4,
DOI:10.1029/2003GC000524, 1-33.
Gorbatov, A., Kostoglodov, V., Suárez, G., and Gordeev, E., 1997. Seismicity
and structure of the Kamchatka subduction zone. Journal of Geophysical
Research, v. 102, no. B8, p. 17,883-17,898.
Graybill, J., Brandon, M.T., and Kepezhinskas, P.K., 1999. Olivine lattivepreffered orientation in xenoliths from the mantle wedge beneath the southern
Kamchatka volcanic arc (abstract). Eos Trans. AGU, 80, F926.
Hamaguchi, H., Goto, K., and Suzuki, Z., 1983. Double-planed structure of
intermediate-depth seismic zone and thermal stress in the descending plate.
Journal of Physics of the Earth, 331, 329-347.
Kelemen, P.B., Rilling, J.L., Parmentier, E.M., Mehl, L., and Hacker, B.R., 2003.
Thermal Structure due to Solid-State Flow in the Mantle Wedge Beneath Arcs: in
Inside the Subduction Factory, Geophysical Monograph 138, edited by Eiler, J.,
AGU, Washington DC, p. 293-311.
- 15 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Kostoglodov, V., Bandy, W., Dominguez, J., and Mena, M., 1996. Gravity and
seismicity over the Guerrero seismic gap, Mexico. Geophysical Research Letters,
23, 3385-3388.
Kostoglodov, V., Singh, S. K., Santiago, J. A., Franco, S. I., Larson, K. M.,
Lowry, A. R., and Bilham, R., 2003. A large silent earthquake in the Guerrero
seismic
gap,
Mexico.
Geophysical
Research
Letters,
30(15),
1807,
doi:10.1029/2003GL017219.
Luhr, J.F., 1997. Extensional tectonics and the diverse primitive volcanic rocks in
the western Mexican Volcanic Belt. Canadian Mineralogy. 35: 473-500.
Manea, M., Manea, V.C., Kostoglodov, V., 2003. Sediment fill of the Middle
America Trench inferred from the gravity anomalies. Geofisica Internacional, 42,
(4), 603-612.
Manea, V.C., Manea, M., Kostoglodov, Currie, C.A., and Sewell, G., 2004a.
Thermal Structure, Coupling and Metamorphism in the Mexican Subduction Zone
beneath Guerrero. Geophysical Journal International, 158, 775–784 doi:
10.1111/j.1365 - 246X.2004.02325.x.
Manea, V.C., Manea, M., Kostoglodov, V., Sewell, G., 2004b. (in press) Thermomechanical model of the mantle wedge in Central Mexican subduction zone and a
blob tracing approach for the magma transport. Physics of the Earth and Planetary
Interiors.
Márquez, A., Oyarzun, R., Doblas, M., Verma, S.P., 1999a. Alkalic (OIB-type)
and calc-alkalic volcanism in Mexican volcanic belt: a case study of plume-related
magmatism and propagating rifting at an active margine? Geology, 27: 51-54.
Márquez, A., Oyarzun, R., Doblas, M., Verma, S.P., 1999b. Replay to comment
to: Alkalic (OIB-type) and calc-alkalic volcanism in Mexican volcanic belt: a case
study of plume-related magmatism and propagating rifting at an active margine?
Geology, 27: 1055-1056.
Márquez, A., Verma, S.P., Anguita, F., Oyarzun, R., Brandle, J.L., 1999c.
Tectonics and volcanism of Sierra Chichinautzin: extension at the front of the
- 16 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
central Trans-Mexican Volcanic Belt. Journal of Volcanology and Geotherm.
Research, 93, 125-150.
Márquez, A. and De Ignatio, C., 2002. Mineralogical and geochemical constrains
for the origin and evolution of magmas in Sierra Chichinautzin, Central Mexican
Volcanic Belt. Lithos, 62, 35-62.
Moore, G., Marone, C., Carmichael, I.S.E., and Renne, P., 1994. Basaltic
volcanism and extension near the intersection of the Sierra Madre volcanic
province and the Mexican Volcanic Belt. GSA Bulletin, 106, 383-394.
Pardo, M., and Suárez, G., 1995. Shape of the subducted Rivera and Cocos
plates in southern mexico: seismic and tectonic implications. Journal of
Geophysical Research, 100, 12357-12373.
Sheth, H., Torres-Alvarado, I., and Verma, S. P., 2000. Beyond subduction and
plumes: A unified tectonic-petrogenetic model for the Mexican Volcanic Belt.
International Geology Review, 42, 1116-1132.
Shubert, D.H., and Cebull, S.E., 1984. Tectonic interpretation of the TransMexican volcanic belt. Tectonophysics, 101, pp.159-165.
Takada, A., 1989. Magma transport and reservoir formation by a system of
propagating cracks. Bulletin of Volcanology, v. 52, pp.118-126.
Takada, A., 1994. The influence of regional stress and magmatic input on styles of
monogenetic and polygenetic volcanism. Journal of Geophysical Research, v. 99,
pp. 13563-13573.
Tatsumi, Y., 1986. Formation of the volcanic front in subduction zones.
Geophysical Research Letters, v. 13, p. 717-720.
van Keken, P.E., Kiefer, B., and Peacock, S.M., 2002. High resolution models of
subduction zones: Implications for mineral dehydration reactions and the transport
of water into deep mantle. G-cubed , 3, 10, 20.
von Hunen, J., van den Berg, A., and Vlaar, N., 2002. On the role of subducting
oceanic plateaus in the development of shallow flat subduction. Tectonophysics,
352, 317-333.
- 17 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Verma, S.P., 1999. Geochemistry of evolved magmas and their relationship to
subduction un-related mafic volcanism at the volcanic front of the central Mexican
Volcanic Belt. Journal of Volcanology and Geothermal Research, v.93, pp. 151171.
Verma, S.P., 2000. Geochemistry of subducting Cocos plate and the origin of
subduction-unrelated mafic volcanism at volcanic front of central Mexican Volcanic
Belt, in Delgado-Granados, H., Aquirre-Diaz, G.J., and Stock, J., eds., Cenozoic
tectonics and volcanism of Mexico: Boulder Co., Geological Society of America,
Special Paper, 334 , pp. 195-222.
Verma S.P., 2002. Absence of Cocos plate subduction-related mafic volcanism in
southern Mexico: A unique case on Earth? Geology, 30, 1095-1098.
Wallace, P., Carmichael, I.S.E., 1999. Quaternary volcanism near the Valley of
Mexico: implications for subduction zone magmatism and the effects of crustal
thickness variations on primitive magma compositions. Contribution to Mineral
Petrology. 135, 291-314.
- 18 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
II. ESTRUCTURA TÉRMICA, ACOPLAMIENTO Y METAMORFISMO EN LA
ZONA DE SUBDUCCIÓN MEXICANA DEBAJO DEL ESTADO DE GUERRRO
Published in: Geophys. J. Int. (2004) 158, 775–784 doi: 10.1111/j.1365246X.2004.02325.x
THERMAL STRUCTURE, COUPLING AND METAMORPHISM IN THE MEXICAN
SUBDUCTION ZONE BENEATH GUERRERO
V.C. Manea1, M. Manea1, V. Kostoglodov1, C.A. Currie2,3, and G. Sewell4
1
Instituto de Geofísica, Universidad Nacional Autónoma de México (UNAM),
México
2
School of Earth and Ocean Sciences, University of Victoria, Victoria, B.C.,
Canada
3
Pacific Geoscience Centre, Geological Survey of Canada, Sidney, B.C., Canada
4
University of Texas, El Paso, USA
ABSTRACT.
Temperature is one of the most important factors that controls the extent
and location of the seismogenic coupled, and transition, partially coupled segments
of the subduction interplate fault. The width of the coupled fault inferred from the
continuous GPS observations for the steady interseismic period and the transient
width of the last slow aseismic slip event (Mw ~ 7.5) that occurred in the Guerrero
subduction zone in 2001 - 2002, extends up to 180 km - 220 km from the trench.
Previous thermal models do not consider this extremely wide coupled interface in
Guerrero subduction zone that is characterized by shallow subhorizontal plate
contact. In this study, a finite element model is applied to examine the temperature
- 19 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
constraints on the width of the coupled area. The numerical scheme solves a
system of 2D Stokes equations and 2D steady state heat transfer equation.
The updip limit of the coupling zone is taken between 100ºC and 150ºC, while the
downdip limit is accepted at 450ºC as the transition from partial coupling to stable
sliding. From the entire coupled zone, the seismogenic zone extends only up to ~
82 km from the trench (inferred from the rupture width of large subduction thrust
earthquakes), corresponding to the 250ºC isotherm. Only a small amount of
frictional heating is needed to fit the intersection of the 450ºC isotherm and the
subducting plate surface at 180 - 205 km from the trench.
The calculated geotherms in the subducting slab and the phase diagram for
MORB are used to estimate the metamorphic sequences within the oceanic
subducting crust. A certain correlation exists between the metamorphic sequences
and the variation of the coupling along the interplate fault.
Keywords: Mexican subduction zone, flat subduction, thermal models, coupling.
- 20 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
INTRODUCTION
The most crucial feature of the Mexican subduction zone is a shallow
subhorizontal plate interface in its central part beneath the Guerrero state
(Kostoglodov et al., 1996). This particular configuration of the young subducting
Cocos plate (~ 14 Myr) apparently creates some distinct geodynamic
consequences, such as very thin continental lithosphere, relatively shallow
intraslab seismicity, remote position of the volcanic front, etc. The Guerrero seismic
gap extending ~ 120 km northwest from Acapulco (Fig. 1) has never ruptured since
1911, meanwhile the neighboring zones suffered large subduction thrust
earthquakes.
Recent continuous GPS observations in Guerrero show that the interplate
coupling during the steady-state interseismic period is abnormally wide, extending
up to 180 - 220 km inland from the trench (Kostoglodov et al., 2003). A few thermal
models of the subduction zone in Guerrero have been proposed (Currie et al.,
2002), however they take no account of the ~ 200 km-wide coupled zone (because
of the lack of this information at that time).
In a recent study of Kostoglodov et al. (2003), the surface deformation
inferred from GPS measurements during this last slow slip earthquake are
compared with the results from the 2D forward dislocation model for an elastic half
space (Savage, 1983). The steady-state component of GPS site velocities is
modeled as constant-velocity slip on the subduction interface. In this approach, a
virtual slip or “back-slip” with magnitude and direction equal and opposite the
relative plate motion, is used to represent frictional coupling on the megathrust. On
any given discrete segment of the megathrust, it is assumed that the steady state
slip rate is some fraction (coupling (α)) of the relative plate motion: α = Sb / S rpm ,
where α = [0 - 1]; Sb is the back-slip rate; Srpm is the relative slip between Cocos
and North American plates (5.5 cm/yr from NUVEL 1A model of DeMets et al.,
1994). When α = 0, no coupling between the two plates is considered (perfectly
- 21 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
decoupled “back-slip” segment) and α = 1 indicates a full coupling between the
oceanic plate and the overriding continental plate (perfectly coupled “back-slip”
segment).
Using the slab geometry from Kostoglodov et al., (1996) and a wide partially
couple zone (~ 220 km from the trench), the dislocation model of Kostoglodov et al.
(2003) show a reasonable well fit with the observations (Fig. 2). The best model
fits the GPS measurements when the plate interface is partially locked on three
segments. The first segment is located in the shallower part of the subducting plate
with a coupling of α = 0.9. It follows two partially coupled segments with α = 0.7.
The rest of the interface slips freely (α = 0.0). In order to obtain a good fit with the
observed data, the partially coupled segment was extended up to 215 km from
trench.
A dislocation model with the slab geometry from Currie et al. (2002) do not
offer a good fit with the observed surface deformations, because of the slab
geometry and the limited extent of the partially coupled zone.
In the view of these new results for the Guerrero subduction zone, the
motivation of the present paper is to review the previous published thermal
structure for Guerrero, and to offer a possible explanation for the largest silent
earthquake ever recorded. The largest slow aseismic slip event in Guerrero (20012002) has developed almost over the entire width of the previously coupled plate
interface (Kostoglodov et al., 2003). The discovery of this large slow aseismic
event and the non-episodic occurrence of such extensive slow earthquake in
Guerrero gap (Fig. 1), call for an examination of the controlling factors and physical
conditions associated with these events.
It is assumed that pressure, temperature and rock composition provide the
key controls on the extent and location of the seismogenic zone (Peacock and
Hyndman, 1999). The main goal of this study is to analyze this wide subhorizontal
coupled plate interface beneath Guerrero using a numerical modeling of the
thermal structure in this subduction zone. The interplate geometry and coupling
extension is better constrained (Kostoglodov et al., 2003) than in the previous
- 22 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
models (Currie et al., 2002). We developed the 2D steady - state thermal models
using a numerical scheme with a system of 2D steady state heat transfer equation
and 2D Stokes equations.
The updip and downdip limits of the interplate and slow slip earthquakes
have been attributed to a certain temperature range. While, the seismogenic,
coupled zone, where large interplate earthquakes often occur, is confined by 100 150ºC and ≤ 350ºC isotherms, the partially coupled, transient zone, is delimited by
≤ 350ºC and 450ºC isotherms (Fig. 3) (Wang, 1980; Tse and Rice, 1986; Blanpied
et al., 1985; Hyndman and Wang, 1993). The position of the updip limit of the
seismogenic zone at 100 - 150ºC has been attributed to the presence of the stable
subducted sliding sediments (Vrolijk, 1990).
Since laboratory experiments (Blanpied et al., 1995) show that continental
rocks exhibit a transition from velocity weakening to velocity strengthening at
325ºC - 350ºC, this temperature range was proposed to be the downdip limit of the
seismogenic zone. However, the experiments were carried out on quartzofeldspathic continental rock type, while the mineralogical composition of the
subducting oceanic crust is quite different. Therefore the 325ºC - 350ºC
temperature range for the dowdip limit of the seismogenic zone should not be
considered so restrictive, it may be quite variable from one subduction zone to
another.
The final intention of this study is to verify a possible relationship between
the predicted metamorphic facies along the subducting oceanic plate (Hacker et
al., 2003) and the width of the interplate coupling inferred from the modeling of the
surface crustal deformations observed during the interseismic steady-state period
and the last silent earthquake in Guerrero.
- 23 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING PROCEDURE
A system of 2D Stokes equations and the 2D steady state heat transfer
equation are solved for the Guerrero cross section (Fig. 1) using the finite element
solver PDE2D (http://pde2d.com/). The equations in an explicit form are:
 
  ∂u ∂v  
∂u 
 ∂  − P + 2η  ∂ η  +  
∂x 
 ∂y ∂x  
 
+ 
=0

∂x
∂y

   ∂u ∂v  

∂v 
 ∂ η  +   ∂  − P + 2η 
∂y 
   ∂y ∂x   + 
=− ρ ⋅g

∂x
∂y

  ∂T
∂T  ∂  ∂T  ∂  ∂T 
 C p  u ∂x + v ∂y  = ∂x  k ∂x  + ∂y  k ∂y  + Q + Qsh





 
(1)
where:
P
- pressure (Pa),
η
- mantle wedge viscosity (isoviscous mantle wedge) (Pa s),
u
- horizontal component of the velocity (m/s),
v
- vertical component of the velocity (m/s),
ρ
- density (kg/m3),
T
- temperature (°C),
Cp
- thermal capacity (MJ/m3 °K),
k
- thermal conductivity (W/m°K),
Q
- radiogenic heat production (W/m3),
Qsh
- volumetric shear heating (W/m3).
Since this paper focuses on the forearc thermal structure, the present
thermal models consider only an isoviscous mantle wedge. Models with strong
- 24 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
temperature-dependent viscosity and magma transport are presented in detail in a
recent paper of Manea et al. (2004).
The Stokes equations are solved only for the mantle wedge, while the heat
transfer equation is solved for the entire model. The linear system solver used by
the present numerical scheme is the frontal method, which represents an out-ofcore version of the band solver (uses a reverse Cuthill - McKee ordering). In the
present numerical scheme, the penalty method formulation is used, P being
 ∂u ∂v 
η
(ε is the
replaced by P = −α '⋅  +  , where α’ is large, on the order of
ε
 ∂x ∂y 
machine relative precision). In other words, the material is taken to be "almost"
incompressible, so that a large pressure results in a small decrease in volume, and
 ∂u ∂v 
the continuity equation  +  = 0 is almost satisfied.
 ∂x ∂y 
The connection between the Stokes and heat transfer equations is the
velocity field. In terms of displacements, the velocity of the oceanic plate is
considered with reference to the continental plate. Thus the convergence rate of
5.5 cm/year between the Cocos and North American plates is used in our models
(DeMets et al., 1994). The velocities in the subducting Cocos slab beneath the
volcanic arc are set of 5.5 cm/yr; therefore the interface with the mantle wedge is
predefined. The boundary between the mantle wedge and overlying lithosphere is
considered fixed.
The finite element grids extend from 20 km seaward of the trench up to 600
km landward. The lower limit of the grid follows the shape of the subducting plate
upper surface (Kostoglodov et al., 1996) at 100 km depth distance. The thickness
of continental crust of 40 km is assumed for the modeling, which is consistent with
the values inferred from the seismic refraction surveys and gravity modeling
(Arzate et al., 1993; Valdes et al., 1986).
The modeled profile is subdivided in three regions: fore-arc, volcanic arc
and back-arc. The continental crust in every region consists of two layers: the
upper crust (15 km thick) and the lower crust (25 km thick). A summary of the
- 25 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
thermal parameters used in the models is presented in Table 1 (compilation from:
Peacock and Wang, 1999; Smith et al., 1979; Ziagos et al., 1985; Vacquier et al.,
1967; Prol-Ledesma et al., 1989).
The average radioactive heat production in the upper continental crust
reported by Ziagos et al. (1985) is about 1.3 ± 0.6 µW/m3. It has an exponential
decrease from the upper crust down to the lower crust. Since the slow slips
occurrence is located in the forearc area, we center the attention to fit the modeled
surface forearc heat flow to the observed heat flow data (Fig. 1). Therefore, the
radioactive heat production for the upper crust is taken of 0.7 µW/m3 (this value is
within the 95% confidence interval of Ziagos et al. 1985’s estimate), while a value
of 0.2 µW/m3 is assumed for the lower crust. This reduction has a negligible effect
on the thermal structure of the subduction interface.
The upper and lower boundaries of the model are maintained at constant
temperatures of 0ºC and 1,450ºC (asthenosphere), correspondingly. The right
(landward) vertical boundary condition (BC) is defined by a 20ºC/km thermal
gradient for the continental crust. This value is in agreement with the back arc
thermal gradient of 17.8 - 20.2ºC/km reported by Ziagos et al. (1985). Although the
conductive heat equation with internal heating does not produce a linear
temperature increase with depth, the heat flow from the mantle controls the thermal
gradient in the crust in the back arc zone. Furthermore, this landward boundary is
located far away (~ 400 km) from the coupled plate interface and does not produce
a significant effect on it (the heat transfer by conduction can be noticeable only to a
relatively small distance). Therefore, we simplify this BC by a linear temperature
increase with depth. It is considered that the temperature at Moho beneath arcs
and backarcs is above 800ºC (Bostock et al., 2002) and 1,450ºC at 100 km depth
in the asthenosphere. In our models we consider the Moho temperature in the
backarc of 850ºC, and a linear thermal gradient for the continental crust of
20ºC/km. Given that in our models the Moho is located at 40 km depth, the mantle
wedge right BC is represented by 10ºC/km thermal gradient down to the depth of
100 km. Underneath 100 km depth no horizontal conductive heat flow is specified.
- 26 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Beneath Moho, for the right boundary corresponding to the mantle wedge, the BCs
are:

 ∂u ∂v  →
∂u  →
 − P + 2 ⋅η ⋅  ⋅ n x + η ⋅  +  ⋅ n y = GB1,
∂x 

 ∂y ∂x 


∂v  →
η ⋅  ∂u + ∂v  ⋅ →
n
P
2
η
+
−
+
⋅
⋅
x

 ⋅ n y = GB 2,
  ∂y ∂x 
∂y 


which are obtained by balancing the internal (stress induced) forces against the
external boundary forces, called tractions (GB1 and GB2). Therefore, beneath
Moho, where there is no "external" force applied, GB1 = GB2 = 0. Since the slab is
considered as a rigid body, for the deepest part of the right boundary, the velocity
of the subducting slab is used. The left (seaward) BC is a one-dimensional
geotherm for the oceanic plate. This geotherm is calculated by allowing a
conductive cooling of the zero age half-space during the time equal to the age of
the oceanic plate at the trench. This geotherm is corrected for the time-dependent
sedimentation history (Wang & Davis, 1992), assuming a constant porosity - depth
profile of the sediment column and a uniform sediment thickness of 200 m (Moore
et al., 1982) at the trench. The calculated oceanic geotherm is shown in Fig. 4.
The plate age at the trench is of 13.7 Myr according to the interpretation of
Pacific-Cocos seafloor spreading magnetic anomaly lineations by Klitgord and
Mammerickx (1982). The plate interface geometry is constrained by the local
seismicity and the gravity anomalies modeling (Kostoglodov et al., 1996). The
Cocos slab has an initial dip of < 15º, which steepens to as much as 35º near the
coast and subsequently becomes subhorizontal beneath the overriding continental
lithosphere. At 270 km from the trench the slab dips into the asthenosphere at 20º
(Fig. 3). The two dense clusters of seismic events beneath the coast (small yellow
circles in Fig. 3), representing the background seismic activity with low magnitude
(Mw ≤ 4), appear to be related with the sharp bending-unbending of the plate in this
region at ~ 80 km and ~ 115 km from the trench. Important stress concentrations
and pressure variations (up to some 100 MPa) along the thrust fault are likely to
appear in this region.
- 27 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Marine heat flow measurements at the Middle American trench (ProlLedesma et al., 1989) revealed anomalously low (Fig. 1) average values of ~ 30
mW/m2, suggesting that the hydrothermal circulation might be active in the upper
part of the oceanic crust near the trench. Unfortunately, the maximum depth as
well as the extension of the hydrothermal circulation layer is unknown. Given all
these uncertainties, our models do not include the cooling of the oceanic plate at
the trench due to hydrothermal circulation. However, the effect of the hydrothermal
circulation becomes insignificant at the distances greater than ~ 100 km from the
trench. The hydrothermal circulation shifts the position of the 100ºC - 150ºC
isotherms with less than 10 km landward (Currie et al., 2002), therefore introducing
only a relatively small error in the estimate of the upper limit of the seismogenic
zone.
The long term continuous sliding between the subducting and the
continental plates along the thrust fault should produce frictional heating. We
introduced in the models a small degree of frictional heating using the Byerlee’s
friction law (Byerlee, 1978). Frictional heating is limited to a maximum depth of 40
km, which corresponds to the contact between the oceanic plate and the mantle
wedge. The pore pressure ratio, PPR (Pore Pressure Ratio, the ratio between the
hydrostatic and lithostatic pressures; PPR ≤ 1; PPR = 1, means no frictional
heating), is set in order to fit the extent of the coupled zone (450ºC isotherm) at
180 km and 205 km from the trench.
The uncertainties in the forearc thermal models arise mainly from errors in
the thermal constants of the continental crust and the oceanic lithosphere and plate
geometry. Underneath the volcanic arc, the major uncertainties come from the
thermal structure of the mantle wedge. Recent thermal models for Central Mexican
Volcanic Belt (Manea et al., 2004), with strong temperature-dependent viscosity,
show an increasing the temperature with < 200ºC along the slab-wedge interface.
A test with reasonably varied parameters show uncertainties in the thermal models
of ±50 - ±100ºC, with the lower limit for the forearc and the higher limit for the
volcanic arc.
- 28 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING RESULTS
The main constraint on the thermal models is the observed surface heat
flow (Fig. 1). The heat flow data show a steep increase in front of the Mexican
volcanic arc that is common for the subduction zones. In our model the fore arc
and the volcanic belt have a surface heat flow of 28 - 35 mW/m2 and ~ 60 mW/m2
respectively. The fore arc heat flow data, 13 - 38 mW/m2, are slightly lower than
those predicted by the thermal models, while for the volcanic arc the measured
values are higher, 64 - 90 mW/m2, than the modeled values. The low modeled heat
flow in the fore arc region of 28 - 35 mW/m2 is a consequence of the heat
consumption from the overriding plate by the underlying cold subducting oceanic
plate. Our models do not consider magma generation and transport, or
temperature-dependent viscosity in mantle wedge; therefore the modeled surface
heat flow beneath the volcanic arc is smaller than the observed values. The
present study focuses only on the fore arc zone, the thermal structure beneath the
volcanic arc does not influence significantly the thrust fault region (Manea et al.,
2004). A recent paper of Manea et al. (2004), shows a better fit of the surface heat
flow in the volcanic arc, due to the introduction of a strong temperaturedependence of the asthenospheric viscosity in the mantle wedge. Rayleigh - Taylor
instabilities may arise at the slab-wedge interface as a consequence of hydration
and partial melting, and compositionally positive buoyant diapirs start to rise toward
the base of the continental lithosphere changing the thermal distribution and flow
pattern in mantle wedge (Gerya and Yuen, 2003). All these effects are not included
in the modeling here.
The examples of thermal models, which correspond to the main restrictions
(e.g., location and extension of the coupled zones, local seismicity and the
hypocenter location of the intraslab earthquakes etc.) are shown in Fig. 5. The
model with PPR = 0.97 is in good agreement with a coupled zone extent up to 180
km from the trench, while the model with PPR = 0.98 better explains the coupled
zone extent up to 220 km. The average shear stress along the thrust fault is 13
- 29 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MPa or an effective coefficient of friction of 0.017 (Wang et al., 1995). A smaller
value of PPR (< 0.97) would increase the amount of frictional heating along the
thrust fault, and the position of the 450ºC isotherm moves toward the trench. A
greater value of PPR (> 0.98) would decrease the frictional heating, and therefore
widening the coupling zone to distances superior to 205 km from the trench. Both
thermal models are in agreement with the hypocenter location of the intraslab
earthquakes (note that the hypocentral depth estimates of the intraplate
earthquakes could have the errors up to 20 km). The models indicate that the
seismogenic fault (limited by 150ºC and 250ºC isotherms) in the Guerrero
subduction zone is in good agreement with the rupture width of large megathrust
earthquakes (Fig. 1) inferred from the aftershock sequences and the models of
seismic rupture (Ortiz et al., 2000). The position of the 450ºC isotherm can account
for the maximum extent of the coupled interplate zone in Guerrero. The slow
aseismic slip events are usually occurring on the transient partially coupled plate
interface limited by the 250ºC and 450ºC isotherms.
- 30 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
METAMORPHIC FACIES IN THE SUBDUCTING SLAB
Interplate coupling in the subduction zone should depend not only on the PT conditions but also on the properties of the rock material at the plate contact. In
our models the pressure is considered hydrostatic, although the non-hydrostatic
stresses and pressures in subduction zones can reach several 100 MPa along the
slab surface. The relationship between the metamorphic facies (Hacker et al.,
2003) and the coupling degree along the subducting oceanic plate can be revealed
using the results of our thermal modeling.
The calculated geotherms of the slab (P-T paths) are plotted on the phase
diagrams for MORB and for harzburgite (Hacker et al., 2003) in order to determine
the metamorphic sequences within the oceanic subducting slab, for our two models
(Fig. 6). The eclogite facies is bounded by lawsonite-blueschist at low
temperatures and by garnet-amphibolite and garnet-granulite facies at high
temperatures. The main metamorphic facies in the Guerrero subduction zone are
shown in Fig. 7. The crustal material of the subducting Cocos plate passes through
zeolite, prehnite-pumpellyite-actinolite facies when T < 250ºC, then it enters into
lawsonite-blueschist-jaedite and epidote-blueschist facies at T < 450ºC. The
position of the hinge point (270 km from the trench) is in good agreement with the
transition to the eclogitic facies.
The maximum depth of the intraslab earthquakes in Guerrero (~ 80 km)
correlates with the depth of the stable hydrous phases suggesting that the
occurrence of these earthquakes might be related with the dehydration process in
the oceanic slab.
The change of metamorphic sequences along the plate interface on the
Guerrero profile is consistent with the location and the extension of the coupled
zones (Fig. 8). The shallow seismogenic zone with α = 0.9, corresponds to the
metamorphic facies of zeolite, prehnite - pumpellyite - actinolite, the intermediate
partially coupled zone with α = 0.7 corresponds to the metamorphic facies of
lawsonite-blueschist, while the deeper zone with α = 0.0 corresponds to the
- 31 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
metamorphic facies of jadeite - lawsonite - blueschist. The blueschist and
associated facies (jadeite and lawsonite) in the range of temperature between
250ºC and 450ºC, and the pressure range of 0.6 - 1.3 GPa, tends to some ductility,
and the slow transient slip events seem to be related with the properties of the
blueschist facies.
The estimated variation of wt% H2O content with depth along the subducting
plate is presented in Fig. 7 - inset. 4 - 5 wt% H2O may be released from the
hydrous phases in the subducting slab through a process of dehydration at the
depths between 20 km and 80 km. Very low shear-wave velocities in the cold
forearc mantle have been discovered in the southern Cascadia subduction zone
(Bostock et al., 2002). This is an evidence of a highly hydrated and serpentinized
material in the forearc region. The same conditions should be expected in the
Guerrero subduction zone, too. The presence of serpentine in the mantle wedge
can be examined using the phase diagram for harzburgite (Hacker et al., 2003).
The calculated geotherms plotted in the phase diagram for harzburgite show that
the serpentine facies might exist in the mantle wedge (Fig. 6-B). The location of
serpentinized mantle wedge tip is critical because it may control the down-dip
coupling and slow slip limits.
- 32 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
DISCUSSION AND CONCLUSIONS
The numerical models of temperature distribution in the forearc for the
central part of Guerrero subduction zone are constrained by: the surface heat flow
data, the shape of the plate interface estimated from gravity modeling (Kostoglodov
et al., 1996), seismicity data and recently estimated extension of the coupled
interplate fault. The modeled seismogenic zone delimited by 100ºC - 150ºC and
250ºC is in good agreement with the rupture width of large subduction thrust
earthquakes in Guerrero. A small degree of frictional heating is required in order to
adjust the downdip extension of the partially coupled zone (450ºC isotherm) in
accordance with the coupled zone width assessed from the deformation models.
The flat subhorizontal plate interface at the distance between 115 and 270 km from
the trench is essential in the model to fit the position of the 450ºC isotherm at ~ 200
km distance from trench corresponding to the maximum extent of the plate
coupling.
The change of the metamorphic sequences in the subducting crust
apparently relates with the variation of the coupling along the interplate fault
estimated from the observations of surface deformation during the interseismic
period (Fig. 8). In the temperature range of 250ºC - 450ºC and the pressure of 0.6 1.3 GPa, the metamorphic facies are represented by jadeite - lawsonite blueschist and epidote - blueschist. The blueschist and associated facies in this
temperature and pressure range probably exposes a ductile behavior, which is
responsible for the long-term partial coupling and the sporadic aseismic transient
slip events. For temperatures superior of 200ºC - 300ºC the preexisting rock
undergoes pronounced changes in texture and mineralogy. Pressure and heat are
the main agents of metamorphism, the effect of the heat on a preexisting rock
(basalt in our case) being the increased ductility and change in mineral
assemblage. Foliated metamorphic rocks, like blueschist, present a layered
texture, favoring the ductile behavior along the slab-overriding plate interface.
Ductile deformation resulting from nonhydrostatic stress, which is characteristic for
- 33 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
subduction zones, is responsible for the development of imposed anisotropic
fabrics in metamorphic rock like blueschist.
The slab surface geotherm of Currie et al. (2002) (Fig. 6) shows a different
path. Although the metamorphic facies along the slab surface are basically the
same, for depths grater than ~ 20 km the P-T limits between adjacent facies are
different. At ~ 250ºC, the slab surface geotherm enters in lawsonite - blueschist,
then between ~ 370ºC and ~ 450ºC the epidote-blueschist facies is present. The
jaedite - lawsonite - blueschist and zoisite - amphibole - eclogite are absent in the
Currie et al. (2002) thermal model for Guerrero; instead jadeite - epidote blueschist, lawsonite - amphibole - eclogite and amphibole - eclogite are present.
Our models show a good correlation between the position of the hinge point (270
km from the trench) and the occurrence of the eclogitic facies in the subducted
oceanic crust (Fig. 7). The thermal models of Currie et al. (2002), illustrate the
initiation of the eclogitic facies at greater depth of ~ 80 km, although the bending of
the slab into the asthenosphere begins at ~ 50 km depth.
The down dip limit of the seismogenic zone proposed by Currie et al. (2002)
extends up to 350ºC. From the thermal models of Currie et al. (2002) and the
phase diagram for MORB, between 250ºC and 350ºC lawsonite-blueschist facies is
present. This temperature range represents half of the total width of the
seismogenic zone. If blueschist and associated facies is responsible for the
occurrence of the slow earthquakes in Guerrero, then it should not be present in
the seismogenic zone. In our models, 250ºC is in good agreement with the
maximum extent of the seimogenic zone (~ 82 km from the trench) and with the
onset of the blueschist and associated facies.
Since the model of Currie et al. (2002) is colder than our models, the
serpentine is not present in the tip of the mantle wedge (Fig. 6-B). The proposed
alternative for the downdip extension of the thrust zone by the presence of the
serpentinezed forearc mantle wedge is not supported by the thermal model for
Guerrero of Currie et al. (2002). Alternatively, the thermal models proposed by this
study reveal the presence of a significant amount of serpentine in the tip of the
- 34 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
mantle wedge (Fig. 6-B and Fig. 7), sustaining the alternative for the downdip
extension of the thrust zone.
According to the phase diagrams for MORB, intensive dehydration in the
subducting oceanic crust should occur (Fig. 7 - inset) at T = 250ºC - 450ºC, and P
= 0.6 - 1.3 GPa, more than 2 wt% H2O being released during this phase
transformation. The occurrence of the slow transient slip events may be related
with this dehydration.
- 35 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
REFERENCES
Arzate, J.A., Mareschal, M. and Urrutia-Fucugauchi, J., 1993. A preliminary
crustal model of the Oaxaca continental margin and subduction zone from
magnetotelluric and gravity measurements. Geofisica Internacional, 32, 441-45.
Blanpied, M.L., Lockner, D.A. and Byerlee, J.D., 1995. Frictional slip of granote
at hydrothermal conditions. Journal of Geophysical Research, 100, 13,045-13,064.
Bostock, M.G., Hyndman, R.D., Rondenay, S., and Peacock, S.M. 2002. An
inverted continental Moho and serpentinization of the forearc mantle. Nature, v.
417, p. 536-538.
Byerlee, J.D., 1978. Friction of rocks. Pure Applied Geophysics, 116, 615-626.
Currie, C.A., Hyndman, R.D., Wang, K., and Kostoglodov, V., 2002. Thermal
models of the Mexico subduction zone: Implications for the megathrust
seismogenic zone. Journal of Geophysical Research, 107, NO. B12, 2370, doi:
10.1029/2001JB000886.
DeMets, C., Gordon, R., Argus, D. and Stein, S., 1994. Effect of recent revisions
to the geomagnetic reversal time scale on estimates of current plate motions.
Geophysical Research Letters, 21, 2191-2194.
Gerya, T.V. and Yuen, D.A., 2003. Rayleigh-Taylor instabilities from hydration and
melting propel ‘cold plumes’ at subduction zones. Earth and Planetary Science
Letters, 212, 47-62.
Hacker, B.R., Abers, G.A., and Peacock, S.M., 2003. Subduction Factory 1.
Theoretical mineralogy, densities, seismic wave speeds, and H2O contents.
Journal of Geophysical Research, v. 108, 10.1029/2001JB001127.
Hyndman, R.D. and Wang, K., 1993. Thermal constrains on the zone of major
thrust earthquake failure: The Cascadia subduction zone. Journal of Geophysical
Research, 98, 2039-2060.
Klitgord, K. and Mammerickx, J., 1982. J., Northern East Pacific Rise: Magnetic
anomaly and bathymetric framework. Journal of Geophysical Research, 87, 67256750.
- 36 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Kostoglodov, V., Bandy, W., Cominguez, J., and Mena, M., 1996. Gravity and
seismicity over the Guerrero seismic gap, Mexico, Geophysical Research Letters,
23, 3385-3388.
Kostoglodov, V., Singh, S. K., Santiago, J. A., Franco, S. I., Larson, K. M.,
Lowry, A. R., and Bilham, R., 2003. A large silent earthquake in the Guerrero
seismic
gap,
Mexico.
Geophysical
Research
Letters,
30(15),
1807,
doi:10.1029/2003GL017219.
Kostoglodov, V., and Pacheco, J. F., 1999. One hundred years of seismicity in
Mexico, Instituto de Geofisica, Universidad Nacional Autónoma de México, México
City.
Manea, V.C., Manea, M., Kostoglodov, V., Sewell, G., 2004 (in press). Thermomechanical model of the mantle wedge in Central Mexican subduction zone and a
blob tracing approach for the magma transport. Physics of the Earth and Planetary
Interiors.
Moore, J.C., Watkins, J.S., Shipley, T.H., McMillen, K.J., Bachman, S.B., and
Lundberg, N., 1982. Geology and tectonic evolution of a juvenile accretionary
terrane along a truncated convergent margin: Synthesis of results from Leg 66 of
the Deep Sea Drilling Project, southern México. Geological Society American
Bulletin, 93, 847-861.
Ortiz, M., Singh, S.K., Kostoglodov, V., and Pacheco, J., 2000. Source areas of
the Acapulco-San Marcos, Mexico earthquakes of 1962 (M 7.1; 7.0) and 1957 (M
7.7), as constrained by tsunami and uplift records. Geofisica Internacional, 39, 337348.
Peacock, S. M., and Hyndman, R. D., 1999. Hydrous minerals in the
mantlewedge and the maximum depth of subduction thrust earthquakes.
Geophysical Research Letters, 26, 2517– 2520.
Peacock, S.M., and Wang, K., 1999. Sesimic consequences of warm versus cool
subduction metamorphism: Examples from southwest and northeast Japan.
Science, 286, 937-939.
- 37 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Prol-Ledesma, R.M., Sugrobov, V.M., Flores, E.L., Juarez, G., Smirnov, Y.B.,
Gorshkov, A.P., Bondarenko, V.G., Rashidov, V.A., Nedopekin, L.N. and
Gavrilov, V.A., 1989. Heat flow variations along the Middle America Trench.
Marine Geophysical Research, 11, 69-76.
Savage, J. C., 1983. A dislocation model of strain accumulation and release at a
subduction zone. Journal of Geophysical Research, 88 , 4984–4996.
Smith, D.L., Nuckels, C.E., Jones, R.L. and Cook, G.A., 1979. Distribution of
heat flow and radioactive heat generation in northern México. Journal of
Geophysical Research, 84, 2371-2379.
Tse, S.T. and Rice, J.R., 1986. Crustal earthquake instability in relation to the
depth variation of frictional slip properties. Journal of Geophysical Research, 91,
9452-9472.
Vacquier, V., Sclater, J.G., and Corry, C.E., 1967. Studies of the thermal state of
Earth. The 21st paper: Heat-flow, eastern Pacific. Bulletin of Earthquake Research
Institute, 45, 375-393.
Valdes, C.M., Mooney, W.D., Singh, S.K., Meyer, R.P., Lomnitz, C, Luetgert,
J.H., Helsley, C.E., Lewis, B.T.R. and Mena, M., 1986. Crustal structure of
Oaxaca, Mexico, form seismic refraction measurements. Bulletin of Seismological
Society of America, 76, 547-563.
Vrolijk, P., 1990. On the mechanical role of smectite in subduction zones.
Geology, 18, 703–707.
Wang, C. Y., 1980. Sediment subduction and frictional sliding in a subduction
zone. Geology, 8, 530– 533.
Wang, K., and Davis, E.E., 1992. Thermal effect of marine sedimentation in
hydrothermally active areas. Geophysical Journal Internacional, 110, 70-78.
Wang, K., Hyndman, R.D. and Yamano, M., 1995. Thermal regime of the
southwest Japan subduction zone: Effects of age history of the subducting plate.
Tectonophysics, 248, 53-69.
- 38 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Ziagos, J.P., D.D. Blackwell, and F. Mooser, 1985. Heat flow in southern Mexico
and the thermal effects of subduction. Journal of Geophysical Research, 90, 54105420.
- 39 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
FIGURE CAPTIONS
Figure 1.
Distribution of the heat flow data and the location of modeled cross-section
in Guerrero. Large orange circles are continental heat flow measurements in
2
mW/m (Ziagos et al., 1985). Small light blue circles in the insets are ocean heat
2
flow measurements in mW/m (Prol-Ledesma et al., 1989). Yellow triangles show
the location of active volcanoes in Mexico. Green squares are the major cities.
Grey thick line is the modeled cross-section. Also, the rupture areas for megathrust
earthquakes along the Mexican coast are shown (Kostoglodov and Pacheco,
1999). The extension of the seismic gap in Guerrero is ~ 120 km northwest from
Acapulco.
Figure 2.
Dislocation model for the interseismic steady state deformations observed
on the GPS stations. The interface is partially locked on three segments (thick
lines; bold numbers indicating the fraction of locking). The rest of the interface slips
freely. The model fits reasonably well the observed data. Displacement errors bars
larger than 1σ are shown. Reproduced from Fig. 4 of Kostoglodov et al. (2003),
copyright by the American Geophysical Union.
Figure 3.
The updip and downdip limits of the seismogenic and slow slip zones. The
seismogenic zone (fine dashed black-pink line) is confined by 100 - 150ºC and ≤
350ºC isotherms. The partially coupled zone, where slow earthquakes are
proposed to occur, (coarse dashed black-pink line) is delimited by ≤ 350ºC and
450ºC isotherms. Beyond 450ºC, the oceanic plate and the overriding plate are
considered completely decoupled (continuous light blue line). Small yellow circles
represent the background seismic activity with low magnitude (Mw ≤ 4). The red
circles represent intraslab earthquakes with magnitude Mw ≥ 5.9.
- 40 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4.
Boundary condition (BC) and thermal parameters used in the modeling. The
upper and lower boundaries are maintained at the constant temperatures of 0ºC
and 1,450ºC (asthenosphere), correspondingly. The right (landward) vertical BC is
defined by 20ºC/km thermal gradient in the continental crust and 10ºC/km up to a
depth of 100 km. Deeper, no horizontal conductive heat flow is specified. The left
(seaward) BC (shown in the inset) is the one-dimensional geotherm for the oceanic
plate. The oceanic geotherm is corrected for the time-dependent sedimentation
history (Wang & Davis, 1992), assuming a constant porosity-depth profile of the
sediment column and a uniform sediment thickness of 200 m at the trench. The
oceanic plate is subducting at a constant rate of 5.5 cm/year while the continental
crust is considered fixed.
The modeled profile is subdivided into three regions: fore-arc, volcanic arc
and back-arc. In each region the continental crust consists of two layers: the upper
crust and the lower crust. A summary of the thermal parameters used in the
models is presented in Table 1.
Figure 5.
(A) Variations of the surface heat flow along the Guerrero profile. Red dots
with vertical error bars are the heat flow measurements from Ziagos et al. (1985).
Blue solid line: the surface heat flow for the model without frictional heating. Red
dashed line corresponds to the model with PPR = 0.97, and the blue solid line to
the model with PPR = 0.97. PPR - Pore Pressure Ratio.
(B) The model of steady-state thermal structures for the 13.7 Myr oceanic
lithosphere subducting at 5.5 cm/year beneath Guerrero. Frictional heating (PPR =
0.98) is introduced down to a maximum depth of 40 km. The shear stress along the
fault is given by Byerlee’s friction law (Byerlee, 1978). Orange triangle Popocatépetl volcano. Black dashed line is the Moho (40 km depth). The
continuous black line indicates the top of the subducting oceanic slab. Shortdashed red segments delimit the seismogenic zone (between 100°C - 150°C and
- 41 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
250°C). Long-dashed pink segment (250°C - 450°C) shows the zone of partial
coupling. The seismogenic zone is located between 32 km and 81 km from the
trench. The coupling zone extends up to 205 km from the trench.
(C) Same as (B) with the frictional heating, PPR = 0.97. The seismogenic
zone is between 32 km and 81 km. The coupling zone extends up to 180 km from
the trench.
Figure 6.
(A). Phase diagram for MORB and maximum H2O contents (Hacker et al.,
2003). Calculated geotherms: continuous blue line and dashed red line - P-T paths
for the top of the subducting oceanic crust for PPR = 0.97 and PPR = 0.98,
respectively; continuous green line - P-T path for the top of the subducting oceanic
slab from Currie et al. (2002). 1 - Zeolite (4.6 wt%H2O), 2 - Prehnite -Pumpellyite
(4.5 wt% H2O), 3 - Pumpellyite - Actinolite (4.4 wt% H2O), 4 - Greenschist (3.3 wt%
H2O), 5 - Lawsonite - Blueschist (5.4 wt% H2O), 6 - Epidote - Blueschist (3.1 wt%
H2O), 7 - Epidote - Amphibolite (2.1 wt% H2O), 8 - Jadeite - Epidote - Blueschist
(3.1 wt% H2O), 9 - Eclogite - Amphibole (2.4 wt% H2O), 10 - Amphibolite (1.3 wt %
H2O), 11 - Garnet - Amphibolite (1.2 wt% H2O), 12 - Granulite (0.5 wt% H2O), 13 Garnet - Granulite (0.0 wt% H2O), 14 - Jaedite - Lawsonite - Blueschist (5.4 wt%
H2O), 15 - Lawsonite - Amphibole - Eclogite (3.0 wt% H2O), 16 - Jaedite Lawsonite - Talc - Schist, (2.0 wt% H2O), 17 - Zoisite -Amphibole - Eclogite (0.7
wt% H2O), 18 - Amphibole - Eclogite (0.6 wt% H2O), 19 - Zoisite - Eclogite (0.3
wt% H2O), 20 - Eclogite (0.1 wt% H2O), 21 - Coesite - Eclogite (0.1 wt% H2O), 22 Diamond - Eclogite (0.1 wt% H2O).
(B) Phase diagram for harzburgite, and maximum H2O contents (Hacker et
al., 2003). A - Serpentine - Chlorite - Brucite (14.6 wt% H2O), B - Serpentine Chlorite - Phase A (12 wt% H2O), C - Serpentine - Chlorite - Dunite (6.2 wt% H2O),
D - Chlorite - Harzburgite (1.4 wt% H2O), E - Talc - Chlorite - Dunite (1.7 wt% H2O),
F - Anthigorite - Chlorite - Dunite (1.7 wt% H2O), G - Spinel - Harzburgite (0.0 wt%
H2O), H - Garnet - Harzburgite (0.0 wt% H2O). Calculated geotherms are the same
- 42 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
as in (A). The calculated geotherms plotted on the phase diagram for harzburgite,
show that in both of our models the serpentine might exist in the mantle wedge.
Figure 7.
Metamorphic facies along the oceanic subducting crust. The metamorphic
facies in the subducting Cocos plate pass through zeolite, prehnite-pumpellyiteactinolite at T ≤ 250ºC then lawsonite-blueschist-jaedite and epidote-blueschist at T
≤ 450ºC. It can be seen from the histogram (inset) wt% H2O versus the
metamorphic sequences along the subducting plate that 4 - 5 wt% H2O may be
released from the hydrous phases in the subducting slab through the process of
dehydration. The presence of the serpentine in the mantle wedge is predicted from
the phase diagram for harzburgite (Hacker et al., 2003) Fig. 6-B. The dashed
yellow line represents the onset of the eclogitic facies in the subducted oceanic
crust.
Figure 8.
Metamorphic facies along the oceanic subducting crust in the forearc.
Superimposed are the coupled segments from the best fitting dislocation model
(Fig. 2). The changes of the metamorphic sequences along the plate interface on
Guerrero profile are consistent with the estimates of the location and the extension
of the coupled zone.
- 43 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Table 1.
Summary of the thermal parameters used in the models. (Compilation from:
Peacock and Wang, 1999; Smith et al., 1979; Ziagos et al., 1985; Vacquier et al.,
1967; Prol-Ledesma et al., 1989).
Thermal Conductivity (W/m°K)
Geological Unit
Fore-
Volcanic-
Back-
arc
arc
arc
1.00 -2.00*
Oceanic sediments
Continental
crust
(0 -15 km)
Continental
crust
(15 - 40 km)
Heat
Thermal
production
Capacity
(µW/m3)
(MJ/m3°K)
1.00
2.50
2.00
3.00
2.50
0.65
2.50
2.00
3.00
2.50
0.20
2.50
Continental mantle
3.10
0.01
3.30
Oceanic lithosphere
2.90
0.02
3.30
* Increase linearly with distance from the deformation front up to a depth of 10 km.
- 44 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 1
- 45 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 2
- 46 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 3
- 47 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4
- 48 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 5
- 49 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 6
- 50 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 7
- 51 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 8
- 52 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
III. MODELO TERMOMECÁNICO DE LA CUÑA DEL MANTO EN LA ZONA DE
SUBDUCCIÓN DEL CENTRO DE MÉXICO Y EL MECANISMO “BLOB
TRACING” PARA EL TRANSPORTE DE MAGMA
In press at: Physics of the Earth and Planetary Interiors
THERMO-MECHANICAL MODEL OF THE MANTLE WEDGE IN CENTRAL
MEXICAN SUBDUCTION ZONE AND A BLOB TRACING APPROACH FOR THE
MAGMA TRANSPORT
V.C. Manea1, M. Manea1, V. Kostoglodov1, G. Sewell2
1
Instituto de Geofisica, Universidad Nacional Autónoma de México (UNAM),
México
2
University of Texas, El Paso
ABSTRACT
The origin of the Central Mexican Volcanic Belt (CMVB) and the influence of
the subducting Cocos plate on the CMVB volcanism are still controversial. In this
study, the temperature and mantle wedge flow models for the Mexican subduction
zone are developed using the finite element method to investigate the thermal
structure below CMVB. The numerical scheme solves a system of 2D Stokes
equations and 2D steady state heat transfer equation.
Two models are considered for the mantle wedge: the first one with an
isoviscous mantle wedge and the second one with a strong temperaturedependent viscosity. The first model reveals a maximum temperature of ~ 830 ºC
in the mantle wedge, which is not sufficient for melting wet peridotite. Also, the
geotherm of the subducting plate upper surface does not intersect the dehydrationmelting solidus for mafic minerals. The second model predicts temperatures of
- 53 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
more than 1,200ºC beneath the CMVB for a wide range of rheological parameters
(reference viscosity and activation energy). Up to 0.6 wt% H2O can be released
down to 60 km depth through metamorphic changes in the oceanic crust of the
subducting slab. The melting of this oceanic crust apparently occurs in a narrow
depth range of 50-60 km and also melting of the hydrated mantle wedge peridotite
is now expected to take place beneath the CMVB.
Considering that the melting processes on and in the vicinity of the
subducting plate surface generate most of the volcanic material, a dynamic model
for the blob tracers is developed using Stokes flow at infinite Prandtl number. The
blobs of 0.2 - 10.0 km in diameter migrate along different trajectories only at low
wrapping viscosities (ηw = 1014 - 5·1017 Pa s). The modeling results show that the
“fast” trajectories terminate at the same focus location at the base of the
continental crust, while the arrival points of “slow” trajectories, which are common
for the blobs of smaller size (~ 0.4 - 0.5 km), are scattered away from the average
focus location. This observation may give us a hint on a possible mechanism of
strato and mono volcanoes genesis. The rise time, which the blob detached from
the subducted plate, needs to reach the bottom of the continental crust, is from
0.001 up to 14 Myr depending on the blob diameter and surrounding viscosity.
Keywords: Mexican subduction zone, thermal models, mantle wedge flow, blobs.
- 54 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
INTRODUCTION
Thermal and flow models in the mantle wedge can give us advance insights
on the geodynamic processes in the forearcs as well as beneath the volcanic arcs.
There are only a few thermal models of the subduction zone in Mexico (Currie et
al., 2002; Manea et al., 2004), but none of them presents a reliable and detailed
thermal structure beneath the volcanic arc considering the rheology of the
asthenosphere. The present models are further developed based on previous
study of the Mexican subduction zone in Guerrero in Manea et al. (2004), which
plausibly constrained the thermal structure for the forearc area only. The thermo mechanical modeling of the mantle wedge with temperature and/or stress
dependent rheology is proposed (Furukawa, 1993; Conder et al., 2002; Van Keken
et al., 2002; Kelemen et al, 2003). Mantle wedge dynamics and thermal structure
of shallow flat subduction zones in general have been investigated recently by van
Hunen et al. (2002). In this study we explore the thermal structure of the mantle
wedge in a specific subduction zone of the Central Mexico (Guerrero), which has
an anomalously wide, subhorizontal plate interface and a distant volcanic arc. The
models describe a stationary slab-induced convection, in the cases of the constant
viscosity (isoviscous mantle) and strong temperature - dependent viscosity of the
asthenosphere.
The previous thermal model for the Guerrero subduction zone with a
shallow plate interface (Currie et al., 2002) with predefined analytical expression
for the mantle corner flow, predicts the temperature of ~ 900ºC in the
asthenosphere beneath the volcanic front (Popocatépetl volcano). In Currie’s
model the basaltic component of the subducting plate crust does not reach the
melting temperature.
Recent geological studies (Luhr, 1997; Márquez et al., 1999a) suggest the
existence of a complex mantle beneath the CMVB. However the origin of the OIBlike basalts is still unclear there. Advection of the asthenospheric mantle caused by
- 55 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
sinking of the Cocos plate was proposed as a possible source of the OIB-like
magmas (Luhr, 1997; Wallace and Carmichael, 1999).
There are several mineralogical and geochemical studies of the CMVB
(Márquez and De Ignacio, 2002) suggesting the existence of two different primitive
mafic magmas, one with an asthenospheric OIB-like component and another with a
lithospheric component. Although the relationship between volcanism and the
subduction of the Cocos plate is commonly recognized (Pardo and Suárez, 1995,
Wallace and Carmichael, 1999), there are different hypothesis regarding the
source of the volcanism in this area: the extension of the Gulf of California
transform fault zone (Ferrari et al., 1994), an old weakened cortical zone (Cebull
and Schubert, 1987), cortical transtension mechanisms (Shubert and Cebull, 1984;
Ferrari et al., 1990) and rifting (Luhr, 1997; Márquez et al., 1999a,b).
The source of fluids that metasomatize the mantle is also unclear. There are
three possibilities proposed to explain the metasomatized mantle (Márquez et al.,
1999c; Wallace and Carmichael, 1999; Verma, 1999; 2000): a) the mantle wedge
affected by the fluids originated from dehydration of the subducted Cocos plate; b)
old enriched lithospheric mantle; c) mantle metasomatized by volatiles from a
mantle plume. Direct evidences that the mantle wedge was metasomatized by
fluids released from the down going oceanic slab is essentially difficult, because
mantle xenoliths are rarely discovered in arc lavas. In the CMVB, Quaternary
hornblende andesites erupted near the El Peñón area (see Fig. 1) contain
xenoliths 1 - 2 cm in diameter. These xenoliths are rich in phenocrysts of
hornblende and the host andesite is depleted in plagioclase phenocrysts (Blatter
and Carmichael, 1998). All these observations suggest an influx of volatiles from
the subducting slab into the mantle wedge.
Underneath the CMVB, the magma ascents toward the earth surface
producing large strato-volcanic structures and smaller sized monogenetic
volcanoes scattered in large areas. The CMVB includes several stratovolcanoes
like Popocatépetl, Iztaccíhuatl and Nevado de Toluca (Fig. 1). The monogenetic
volcanoes are basically cinder cones, lava cones, domes and lava flows. While
- 56 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
stratovolcanoes are characterized by a periodicity of magma eruptions, the
monogenetic volcanoes present a single one eruptive event, and as a result have a
smaller size compared with stratovolcanoes. Fedotov (1981) suggested that the
occurrence of one or another type of volcanoes might be related with the magma
supply. Other scientists (Takada, 1989, 1994) propose a dual mechanism related
with both the magma supply and regional stress. While the alignment of the
monogenetic volcanoes can be parallel to a main normal fault system, the
arrangement of the large stratovolcanoes seems to be rather orthogonal to the
volcanic belt (Alaniz-Alvarez et al., 1998).
Using the phase diagrams for mafics (e.g. Hacker et al., 2002) it is possible
to investigate whether the mantle wedge is subjected to the hydration by fluids
released from the subducting slab. A very thin sedimentary fill at the Mexican
trench suggests that about of 95% of these sediments (~ 200 m) are subducted
(Manea et al., 2003). An influx of volatiles from the metamorphosed oceanic crust
and sediments might trigger partial melting of the peridotite just above the
subducted slab (Tatsumi, 1986; Davies and Stevenson, 1992). Gerya and Yuen
(2003) showed that a Rayleigh-Taylor instability developed above the subducting
slab generates positively buoyant plumes up to 10 km in diameter that can
penetrate the overlying mantle wedge.
Albeit various mechanisms of magma generation have been proposed (e.g.
anhydrous decompression melting of peridotite (Klein and Langmuir, 1987;
Langmuir et al., 1992); porous flow of hydrated partial melt (Davies and Stevenson,
1992)), in the present study, the magma generation and migration is assumed in a
form
of
partially
melted
positively
buoyant
blobs.
Regardless
of
the
oversimplification of the blob properties, this model can help to understand the
existence of different sources of the volcanism in the area. The buoyant blobs of
different size and composition may be generated by melting of the Cocos plate and
overlying mantle peridotite, when the pressure and temperature reach the solidus
conditions (e.g. Gerya and Yuen, 2003).
- 57 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
It is important to estimate the viscosity range for the reasonable trajectories
and average rise times for the blobs reaching the bottom of the continental crust. A
recent study by Gerya and Yuen (2003) revealed that these plume - like blobs may
be lubricated by the partially melted material of the subducted crust and hydrated
mantle, thus producing a very low viscosity wraps around the blob structures.
Burov et al. (2000), apply a similar extremely low viscosity, in order to model the
exhumation in the continental lithosphere.
- 58 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING PROCEDURE
A system of 2D Stokes equations and 2D steady state heat transfer
equation is solved for the Guerrero cross section (Fig. 1) using the finite element
solver PDE2D (http://pde2d.com/). The system of equations in an explicit form is:
 
  ∂u ∂v  
∂u 
 ∂  − P + 2η  ∂ η  +  
∂x 
 ∂y ∂x  
 
+ 
=0

∂x
∂y

   ∂u ∂v  

∂v 
 ∂ η  +   ∂  − P + 2η 
∂y 
   ∂y ∂x   + 
= − ρ ⋅ g − Ra ⋅ T

∂x
∂y

  ∂T
∂T  ∂  ∂T  ∂  ∂T 
 C p  u ∂x + v ∂y  = ∂x  k ∂x  + ∂y  k ∂y  + Q + Qsh





 
where:
P - pressure (Pa),
η = η0 ⋅ e
 Ea  T0  
⋅  −1  

 R ⋅T0  T  
- the mantle wedge viscosity (Pa s).
Other parameters are:
η0
-mantle wedge viscosity at the potential temperature T0 (reference
viscosity) (1017 -1021 Pa s),
T0
-mantle wedge potential temperature (1,450ºC),
Ea
-activation energy for olivine (kJ/mol),
R
-universal gas constant (8.31451 J/mol.K),
T
-temperature (ºC),
u
-horizontal component of the velocity (m/s),
v
-vertical component of the velocity (m/s),
ρ
-density (kg/m3),
- 59 -
(1)
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Cp
-thermal capacity (MJ/m3K),
k
-thermal conductivity (MJ/m3K),
Q
-radiogenic heating (W/m3),
Qsh
-volumetric shear heating (W/m3),
ρ ⋅α ⋅ g ⋅ ∆T ⋅ L3
Ra =
η0 ⋅ k
-thermal Rayleigh number,
α
-thermal expansion 3.5·10-5 (1/ºC),
L
-length scale (330 km),
∆T
-temperature difference between the bottom and top model temperatures
(1450 ºC),
k
-thermal diffusivity (10-6 m/s2),
g
-gravitational acceleration (9.81 m/s2).
The Stokes equations are solved only for the mantle wedge, while the heat
transfer equation is solved for the entire model. The linear system solver used by
the present numerical scheme is the frontal method, which represents an out-ofcore version of the band solver (uses a reverse Cuthill - McKee ordering). For the
model with strong temperature-dependent viscosity, the system of equations
becomes strongly nonlinear, therefore Picard iterations are applied, and in order to
achieve a convergent solution a cut-off viscosity of 1024 Pa s for the temperature
less than 1,100 ºC is used. For such highly non-linear problems we constructed a
one-parameter family of problems using a variable (V), such that for V = 1 the
problems is easy (e.g. linear) and for V > N (N is less NSTEPS = 20), the problem
reduces to the original highly nonlinear problem. The nonlinear terms are multiplied
by MIN(1.0,(V - 1.0)/N). With a wide viscosity range from 1017 Pa s to 1024 Pa s,
the fully nonlinear viscosity formulation (e.g., temperature and stress dependence
of the viscosity) presented significant numerical instabilities; therefore the strainrate dependence is neglected in the present study.
- 60 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
In the present numerical scheme, the penalty method formulation is used, P
 ∂u ∂v 
η
being replaced by P = −α '⋅  +  , where α’ is large, on the order of
(ε is
ε
 ∂x ∂y 
the machine relative precision). In other words, the material is taken to be "almost"
incompressible, so that a large pressure results in a small decrease in volume, and
 ∂u ∂v 
the continuity equation  +  = 0 is almost satisfied.
 ∂x ∂y 
The finite elements grid extends from 25 km seaward of the trench up to 600
km landward of it, and consists of 12,000 triangular elements with the higher
resolution in the tip of the wedge (Fig. 2). The triangles have the height to width
ratio equal to 1. A benchmark with different grid resolutions is done to quantify the
numerical error introduced by the present numerical scheme. The lower limit of the
grid follows the shape of the subducting plate upper surface (Kostoglodov et al.,
1996) at 100 km depth distance. The top of the model has a fixed temperature of
0°C. The temperature at the bottom of the model is of 1,450°C which represents
the mantle temperature at ~ 100 km depth (Fig. 3).
The continental lithosphere in Guerrero is defined mostly by the crust, which
thickness is 40 km in the model. This is consistent with the values inferred from the
seismic refraction surveys and gravity modeling (e.g. Valdes et al., 1986, Arzate et
al., 1993). The bending geometry of the subducting slab (up to the hinge point: 270
km) is well constrained by gravity modeling (Kostoglodov et al., 1996), seismicity
data and recently estimated extension of the coupled plate interface inferred from
GPS measurements during the last slow slip event in Guerrero (2001-2002)
(Kostoglodov et al., 2003).
A dip of the subducting plate beneath the volcanic arc is poorly constrained
because of a very limited number of intraslab earthquakes. The dip of 20º and a
hinge point at 270 km from the trench match better the hypocenter locations of the
intraslab events (Fig. 4) and as a result, these are enveloped by the modeled
“seismicity cut-off” temperature of T ~ 800°C (Gorbatov and Kostoglodov, 1997).
The continental crust consists of two layers: the upper crust of 15 km and the lower
- 61 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
crust of 25 km. A summary of the thermal parameters used in the models is
presented in Table 1 (compilation from: Peacock and Wang, 1999; Smith et al.,
1979; Ziagos et al., 1985; Vacquier et al., 1967; Prol-Ledesma et al., 1989).
The radiogenic heat generation in the continental crust decreases
exponentially from 1.3 µW/m3 in the uppermost crust down to 0.2 µW/m3 in its
lowermost part (Ziagos et al., 1985). The radiogenic heat production in the models
is taken as 50% from the above values to fit the surface heat flow with the
observed heat flow data (Fig. 5). This reduction has a minor effect on the thermal
structure of the subduction interface and mantle wedge.
A long term sliding between the subducting and the continental plates along
the thrust fault should produce frictional heating. We introduced in the models a
small degree of volumetric frictional heating using the Byerlee’s friction law
(Byerlee, 1978). Frictional heating is ceased at a maximum depth of 40 km, which
corresponds to the contact between the oceanic plate and the mantle wedge (Fig.
3). The volumetric shear heating is calculated as follows:
Qsh =
τ ⋅v
w
(2)
,
where:
Qsh
-volumetric shear heating (mW/m3),

τ = 0.85 ⋅ σ n ⋅ (1 − λ ) for σ n ⋅ (1 − λ ) ≤ 200MPa

τ = 50 + 0.6 ⋅ σ n ⋅ (1 − λ ) for σ n ⋅ (1 − λ ) > 200MPa
τ
-shear stress
σn
-lithostatic pressure (MPa),
λ
-pore pressure ratio, (the ratio between the hydrostatic and lithostatic
pressures. λ = 0.98 in the present study. The maximum value, λ = 1, means
no frictional heating),
v
-convergence velocity (5.5 cm/yr),
w
-thickness of the oceanic crust involved in friction (200 m).
- 62 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The right landward vertical boundary condition is defined by 20ºC/km
thermal gradient in the continental crust. This value is in agreement with the backarc thermal gradient of 17.8 - 20.2ºC/km reported by Ziagos et al. (1985).
Underneath Moho (40 km), the right boundary condition is represented by
10ºC/km thermal gradient down to the depth of 100 km. Below 100 km, no
horizontal conductive heat flow is specified. Below 100 km, no horizontal
conductive heat flow is specified. Beneath Moho (40 km depth), for the right
boundary corresponding to the mantle wedge, the boundary conditions are:

 ∂u ∂v  →
∂u  →
2
η
η
P
n
−
+
⋅
⋅
⋅
+
⋅
x

 +  ⋅ n y = GB1,

∂x 

 ∂y ∂x 


∂v  →
η ⋅  ∂u + ∂v  ⋅ →
n x +  − P + 2 ⋅η ⋅  ⋅ n y = GB 2,


  ∂y ∂x 
∂y 


(3)
which are obtained by balancing the internal (stress induced) forces against the
external boundary forces, called tractions (GB1 and GB2). Therefore, beneath
Moho, where there is no "external" force applied, GB1 = GB2 = 0.
At the intersection between the subducted slab and the right boundary, the
velocity of the subducting slab is used.
The left seaward boundary condition is a one-dimensional geotherm
calculated for the oceanic plate by allowing a half-space to cool from zero age to
the oceanic plate age at the trench. This geotherm is obtained using a timedependent sedimentation history (Wang and Davis, 1992) and assuming a
constant porosity-depth profile of the sediment column with a uniform sediment
thickness of 200 m (Moore et al., 1982) at the trench (Fig. 3). The sedimentation
history and the porosity-depth profile are used only to calculate the oceanic
geotherm and are not included in the modeling procedure.
In terms of displacements, the velocity of the oceanic plate is considered
with reference to the continental plate. Thus the convergence rate of 5.5 cm/yr
between the Cocos and North American plates is used for the Guerrero subduction
- 63 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
zone (DeMets et al., 1994). The velocities in the subducting Cocos slab are set of
5.5 cm/yr; therefore the interface with the mantle wedge is predefined. The
boundary between the mantle wedge and overlying lithosphere is considered fixed.
The Cocos plate age at the trench is 13.7 Myr according to the interpretation
of Pacific-Cocos seafloor spreading magnetic anomaly lineations (Klitgord and
Mammerickx, 1982, Kostoglodov and Bandy, 1995). The forearc thermal model
constraints are described in detail by Manea et al., 2004. The present study is
focusing mostly on the mantle wedge thermal and velocity structure.
Based on the velocity field obtained in the case of temperature dependent
viscosity, a dynamic model for the blob tracers is developed using Stokes flow at
infinite Prandtl number (Turcotte and Schubert, 1982). The blob moves under the
action of drag, mass, and buoyancy forces in the mantle wedge stationary velocity
field generated in the previous model (1). To investigate the mere dynamic effect of
the mantle wedge convection on the hypothetical blobs rising from the subducted
plate up to the base of the continental lithosphere, the following main assumptions
are done:
-The motion of the asthenosphere is described by the above 2D Stokes equations
(1);
-The blobs are spherical and the drag force is assumed to be similar to that of nondeforming spheres;
-The velocity field is steady state and the liquid acceleration is negligible because
the related forces have small magnitude compared to the steady drag force;
-The influence of the blob motion on the mantle wedge circulation is irrelevant and
there is not interaction between individual blobs.
According to these assumptions the total force acting on the blob is:
(4)
,
- 64 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
where:
→
4 ⋅π ⋅ r 3
d vb
F=
⋅ ρb ⋅
dt
3
→
-resultant force
→
4
⋅ π ⋅ r 3 ⋅ ρb ⋅ g
3
→
→
4
F A = − ⋅π ⋅ r 3 ⋅ ρa ⋅ g
3
→
→
→
→
→
1
F D = ⋅ π ⋅ r 2 ⋅ ρ a ⋅ CD (Re) ⋅ (ν a − ν b )⋅ | ν a −ν b |
2
24
CD (Re) =
Re
→
Fg =
→
→
-steady drag force,
-drag force coefficient
→
2 ⋅ r⋅ | ν a −ν b |
Re =
ηw / ρ a
-Reynolds number,
→
-blob velocity,
→
va
-mantle wedge velocity,
r
-blob radius (0.1 - 5.0 km),
ρb
-blob density (3,000 kg/m3),
ρa
-mantle wedge density (3,200 kg/m3),
→
g
-gravitational acceleration (9.81 m/s2),
ηw
-the blob’s wrapping viscosity (Pa s).
vb
→
-buoyancy force ( F A − F g ),
Using the expressions for forces, the equation of the blob motion has the
following form:
→
→
→
d v b → ρ a → 3 ρ a CD ( Re ) → →
g+
=g−
( v a − v b )⋅ | v a − v b |
dt
ρb
8 r ρb
- 65 -
(5)
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The equations of the blob motion can be written as a system of four ordinary
differential equations, two equations for the coordinates and two for the velocity.
Then the system is solved using PDE2D.
The time step size will be chosen adaptively, between an upper limit of
DTMAX = TF/NSTEPS and a lower limit of 0.0001·DTMAX. NSTEPS represents
the minimum number of steps (NSTEPS = 5) and TF represents the time
necessary for a blob to touch the base of the continental crust. Each time step, two
steps of size DT/2 are taken, and that solution is compared with the result when
one step of size DT is taken. If the maximum difference between the two answers
is less than the relative tolerance (0.001), the time step DT is accepted. Then, the
next step DT is doubled, if the agreement is "too" good; otherwise DT is halved and
the process is repeated. As the tolerance is decreased, the global error decreases
correspondingly. The Crank-Nicolson scheme is used to discretize the time. The
→
steady mantle wedge velocity ( va ) field is obtained previously, for the model with
temperature dependent viscosity (1).
The trajectories of blobs with the diameters varying between 0.2 and 10.0 km are
calculated for different values of ηw. The total rise times that the blobs require to
reach the base of the continental crust are also estimated.
- 66 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING RESULTS
The thermal models which, correspond to the isoviscous mantle wedge and
to the temperature-dependent viscosity, are presented in Fig. 4, Fig. 6 and Fig. 7.
The isoviscous mantle wedge model predicts temperatures of ~ 830ºC in the
asthenosphere (Fig. 4), beneath the Popocatépetl volcano, indicating that melting
should not occur at least for dry olivine. The geotherm of the oceanic plate surface
does not intersect the solidus for basalt (Fig. 8-A), suggesting that the oceanic
plate does not suffer melting as well. The temperature at the base of the
continental crust is also low, reaching values of ~ 800ºC, which is not sufficient to
produce melting. The back flow velocity field from the mantle wedge (the inflow into
the mantle wedge) is relatively small, ~ 1.5 cm/yr. It is evident that simple model
with the isoviscous mantle wedge cannot create any source of the volcanic
material beneath the CMVB.
A series of benchmark tests with different grid resolutions have been done
to verify the accuracy of numerical scheme realized in the present study. The slabmantle wedge interface is the area where the isotherms have a very small
incidence angle with this interface, thus significant errors might appear along this
boundary. The benchmark tests on this boundary quantify the numerical errors as
a function of mesh resolution. Grids with a systematic increase in element number
from 4,000 to 12,000 elements in steps of 2,000 elements are used for this
benchmark.
The benchmark results are presented in Fig. 9. Large temperature
fluctuations (up to 22ºC) close to the tip of the wedge occur for grids with 4,000 and
6,000 elements. Increasing the mesh resolution (8,000 and 10,000 triangles), this
fluctuation are diminished and the maximum temperature fluctuation along the
slab-wedge interface is of ~ 7ºC. Increasing the grid resolution to 12,000 triangles,
the numerical error is less then 5ºC. We stop the benchmark at this point because
the computing time is increasing exponentially with the grid resolution (from half of
hour for 4,000 triangles to more than 10 hours for a model with 12,000 elements,
- 67 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
running on a Pentium 4 PC at 3 Ghz and 2 Gb RAM memory). To obtain final
results we used the 12,000 elements grid, thus our thermal models have the
numerical error < 5ºC.
The temperature dependent viscosity case has been investigated for a
systematic variation of the rheological parameters η0 and Ea. Hirth and Kohlstedt
(2003) showed that the viscosities experimentally estimated for olivine at upper
mantle pressures and temperatures are in the range of 1017 - 1021 Pa s.
Consequently, the reference viscosity, η0, has been varied in our model from 1017
Pa s up to 1021 Pa s. The modeling results are presented in Fig. 6. The activation
energy for diffusion creep in olivine is fixed at 300 kJ/mol (Karato and Wu, 1993).
For the viscosity, η0, in the range from 1017 Pa s to 1020 Pa s, only a small increase
of temperature (< 1 ºC) is observed beneath Popocatépetl (Fig. 10). The maximum
temperature underneath the Popocatépetl is around 1,260 ºC. On the other hand,
for the reference viscosity of η0 = 1021 Pa s, a significant decrease of temperature
(~ 200ºC) occurs. For this case a maximum temperature below Popocatépetl is ~
1,170ºC. The viscosity distributions in the mantle wedge for η0 = (1017 - 1021 Pa s)
are presented in Fig. 11.
Experimentally obtained values of the activation energy for diffusion creep in
olivine are of 300 kJ/mol (Karato and Wu, 1993) and of 315 kJ/mol (Hirth and
Kohlstedt, 1995). The thermal models with variable activation energy from 150
kJ/mol up to 350 kJ/mol are presented in Fig. 7. Since the variation of the
reference viscosity has a little effect on the overall thermal distribution (except for
η0 = 1021 Pa s), the modeling is done for constant η0 =1020 Pa s. Increase of the
activation energy from 150 kJ/mol up to 300 kJ/mol results in a relatively small
increase of temperature (< 25ºC). For higher activation energies (~ 350 kJ/mol) an
important increase in temperature (up to 70ºC) is observed, and the maximum
temperature below the Popocatépetl volcano reaches ~ 1,330 ºC (Fig. 12). The
mantle wedge viscosity distributions for Ea = (150 - 350 kJ/mol) are presented in
Fig. 13.
- 68 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The geotherms along the slab surface are presented in the phase diagram
for mafics and harzburgite (Hacker et al., 2002) for a variable reference viscosity,
η0, in Fig. 8, and for variable activation energy, Ea, in Fig. 14. The slab geotherm
intersects the dehydration melting solidus for basalt at 50 - 60 km for the reference
viscosities η0 =(1017 - 1020 Pa s) and the activation energy Ea = (150 - 300 kJ/mol)
(Fig. 8-A and Fig. 14-A).
The effect of the mantle wedge flow on the blob’s trajectory is an interesting
aspect of the volcanic magma source problem. We selected an initial point for the
blob trajectories at the depth of 70 km on the surface of the subducted slab. This
initial point corresponds to the geometric vertical projection of the Popocatépetl
volcano on the subducting Cocos plate. Estimated blob trajectories are presented
in Fig. 15 for different wrapping viscosities (1014 - 5·1017 Pa s) and blob diameters
(0.2 - 10.0 km).
For a wide range of activation energies (150 - 350 kJ/mol) and reference
viscosities (1017 - 1020 Pa s), the blob trajectories and rising times are practically
indistinguishable (Fig. 16). Only for the reference viscosity of 1021 Pa s, an ~ 1 Myr
increase in rising time is obtained due to lower velocities in the mantle wedge. A
very low wrapping viscosity is essential to let the blob to rise up to the continental
crust. Such low viscosity could result from the lubricating wrap around the blob.
The source of the wrapping is apparently the melted material coming from the
subducting slab, including the melted subducted sediments. Indeed, the water
saturated sediments are likely to melt at a depths of 50 - 58 km for η0 =(1017 - 1021
Pa·s) (Fig. 17-A), and of 45 - 50 km for Ea = (150 - 350 kJ/mol) (Fig. 17-B).
For the blobs of 2 km size (Fig. 15-C) and the wrapping viscosity, ηw >
2·1016 Pa s, the drag force is predominant and the blob cannot rise. Decreasing the
viscosity the drag force is less significant and at the depth of ~ 110 km the blob
intercepts the mantle wedge back flow, which returns it toward the tip of the wedge.
Finally the blob rises up and touches the continental crust after ~ 8 Myr. For the
lower viscosity (< 9·1015 Pa s) the blobs are rising faster (Fig. 16) and pop up at
approximately the same point below the Popocatépetl volcano (~ 350 km from the
- 69 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
trench). The larger is the blob’s size the less time is necessary to reach the
continental crust. Any blob of the size > 0.6 km will touch the bottom of the
continental crust in less then 4 Myr when the wrapping viscosity is of 1·1015 Pa s
(Fig. 15-A). For the blob with the diameter of ~ 2 km, ~ 1 Myr is necessary to pop
up if the ηw < 3·1015 Pa s. The buoyancy force of large size blobs becomes more
dominant than the drag force and this yields a substantially upright trajectory. A 10
km blob touches the Moho in almost the same location for ηw < 1017 Pa s (Fig. 15B). On the other hand, 10 km blobs with the ηw > 5·1017 Pa s would never rise up to
the continental crust (Fig. 15-D). For a ηw fixed at 1017 Pa s, blobs with diameters
grater than 4 km will be able to traverse the mantle wedge flow and to go up
toward the surface.
- 70 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
DISCUSSION AND CONCLUSIONS
Numerical models of temperature and velocity fields in the mantle wedge
can contribute to our understanding of the magma generation, dynamics and
volcanic sources in very atypical subduction zone of Central Mexico. In particular,
the models, allowing for the ascending of buoyant material melted from the
subducted slab and mantle in the form of plumes or blobs, can reveal the
mechanisms of magma transport to the bottom of the continental plate.
Two basic thermo-mechanical models for the mantle wedge in the Central
Mexico are considered in this study. The first model is restricted with the isoviscous
mantle wedge whereas the second one is advanced with the temperaturedependent mantle viscosity. The first model (Fig. 4) does not predict any melting
conditions in the asthenosphere, beneath the CMVB, at the base of the continental
crust and on the surface of the subducting slab. The slab surface geotherm
intersect neither the dehydration melting solidus for basalt (Fig. 8-A) nor the solidus
for H2O saturated sediments (Fig. 17-A). The vertical temperature profile just
beneath the Popocatépetl volcano through the mantle wedge does not encounter
the wet peridotite solidus (Fig. 8-B), thus the melting of the hydrated peridotite
should not occur.
The temperature and velocity fields in the second model depend on the
rheological parameters, the reference viscosity, η0, and the activation energy, Ea.
Variation of the reference viscosity, η0, from 1017 Pa s to 1020 Pa s causes a slight
temperature increase (< 15ºC) close to the tip of the wedge (Fig. 10). Increasing
the reference viscosity up to η0 = 1021 Pa s provokes a significant drop of the tip
temperature of about 200ºC. The phase diagram for mafic minerals shows (Fig. 8A) that the melting of basaltic oceanic crust might take place at the depths of ~ 58
km for η0 = (1017 - 1020 Pa s), and at ~ 80 km for η0 = 1021 Pa s. The water
saturated sediments start to melt at shallower depths, between 50 km and 55 km
(Fig. 17-A). The vertical temperature profile beneath the Popocatépetl volcano
achieves the temperature which is high enough to melt the wet peridotite (Fig. 8-B).
- 71 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The thermal models show that basalt melting initiates at ~ 60 km depth for
the activation energy, Ea = 150 - 300 kJ/mol, and at ~ 50 km depth for Ea = 350
kJ/mol (Fig. 14-A). In the range of reference viscosity values (1017 - 1021 Pa s), the
saturated sediments may melt at the depth between 45 km and 50 km (Fig. 14-B).
The temperature below the Popocatépetl volcano is well beyond the wet peridotite
solidus for the whole range of activation energies applied in this study (150 - 350
kJ/mol).
The modeling results ascertain that the melting of the oceanic crust is likely
to occur in a narrow depth range of 50 - 60 km (except for a reference viscosity of
1021 Pa s which predicts the melting at ~ 80 km depth). The subducted sediments
begin to melt at shallower depths of 45 - 55 km for the entire range of η0 and Ea.
As the oceanic crust melting starts, at first, the dacitic-rhyolitic magma is
formed; afterward as it ascends and interacts with the mantle wedge, the adakitic
magma might be formed. The oceanic slab contribution to the volcanism in Eastern
Mexican Volcanic Belt has been reported by Gomez-Tuena et al. (2003). Magmatic
rocks with the adakitical signature have been found recently in the Quaternary
series in CMVB. The stratovolcano Nevado de Toluca shows evidences of the
same adakitic mark as well (Gomez-Tuena - personal communication).
The temperature of the hydrated peridotite below the CMVB is beyond wet
peridotite solidus, but it is still lower the dry solidus. This suggests that the
hydration of the mantle wedge by fluids released from the subducted Cocos plate
is a necessary condition for the partial melting of the mantle. Indeed, the
metamorphic dehydration predicted by our thermal models might occur down to 80
km depth. The estimated variation of wt% H2O content with the depth along the
subducting plate is presented in Fig. 8-A - inset. By the transformation of zoisite
and amphibole into eclogite, ~ 0.6 wt% H2O may be released into the mantle
wedge from the hydrous phases in the subducting slab through a dehydration at
the depths between 40 km and 60 km.
Mantle xenoliths (oxidized peridotite) have been found in Mexico near El
Peñón (Fig. 1), suggesting an important flux of volatiles from the subducting Cocos
- 72 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
slab (Blatter and Carmichael, 1998). This dehydration would drop down the mantle
viscosity in the tip of the wedge. This effect is not included in the present thermal
models. The amount of serpentine in the tip of the wedge is much smaller for the
model with temperature dependent viscosity (Fig. 8-B -upper left inset) than for the
model with the isoviscous mantle wedge (Fig. 8-B -upper right inset). Furthermore,
the amount of serpentine decreases as the activation energy increases (Fig. 14-B upper insets). The serpentinized tip of the mantle wedge may have bearing on the
down dip limit of the forarc coupled plate interface that controls an extension of
aseismic slow slip transients (Kostoglodov et al., 2003; Manea et al., 2004).
The calcalkaline rocks are the most common series in the CMVB. This
series includes the igneous rocks from basalts up to rhyolites, and is characterized
by depletion of Fe. The depletion probably occurs because of the Fe and Ti oxides
crystallization, which initially might be facilitated by the presence of fluids in the
magma. The influx of volatiles from the metamorphosed oceanic crust and
sediments triggers partial melting of peridotite above the subducted slab (Tatsumi,
1986; Davies and Stevenson, 1992). The majority of the calcalkalines in the CMVB
is represented by the rocks with high content of K2O and Na2O, which originally
might represent low degrees of partial melting in the mantle wedge. In fact, the wet
solidus conditions for peridotite (Wyllie, 1979) are developed close to the slab
surface (see Fig. 7-A).
Felsic magma formations were found in the monogenetic volcanic field of
Chichinautzin (Márquez and De Ignacio, 2002). The models with the temperaturedependent viscosity show that at the base of the continental crust, the temperature
exceeds 1,100ºC (Fig. 6 and Fig. 7). That may create felsic magma sources in
Sierra Chichinautzin by partial melting of the basaltic lower crust under low water
fugacity conditions.
Recent paper of Gerya and Yuen (2003) demonstrates that Rayleigh-Taylor
instabilities can develop and rise up from the surface of the cold subducting slab.
They also suggest that the plumes detached from the slab might be lubricated by
partially melted, low viscosity material of the subducted crust and hydrated mantle.
- 73 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The modeling of the blob motion in the mantle wedge viscous flow induced
by the subducting slab shows that this simple approach may help to understand
the origin of the volcanism in the CMVB. The plume like blobs might origin at the
slab-mantle wedge interface as a consequence of thermal instability. It appears
that the detachment area along the subducted slab from which the blobs might
emerge is not very large. Within the wide range of values of the rheological
parameters, the sediment and oceanic crust melting, and metamorphic dehydration
is expected to occur in the present models at depths of 45 - 60 km.
A certain distance along the slab surface is necessary for the melted
material to accumulate and to form the blob. Nevertheless, the same blob
detachment point in the present model is selected on the subducted slab surface,
just below the main volcanic structure, Popocatépetl volcano (70 km depth). In
reality, the spatial (and temporal) variation of the composition of the material
transferred from the subducted Cocos plate into the overlying mantle wedge is
expected. Therefore the volcanic arc lavas may be enriched with incompatible
elements and volatiles, and highly variable composition of the subducted-related
lavas is probable. The proposed positively buoyant blobs might have a complex
composition of melted H2O saturated peridotite, melted sediments and oceanic
crust.
Two parameters control the trajectories of the blob structures rising from the
slab: the diameter of the blob and the wrapping viscosity. Very low values of the
wrapping viscosity (1014 - 5·1017 Pa s) are necessary to reduce the drag force,
which is critical for the blob to pop up. The lowest dynamic viscosity reported for
dry mantle is ~ 1017 Pa s (Moore et al., 1998). On the other hand, the viscosity of
hydrated, partially molten blobs might be as low as 1014 Pa s (Gerya and Yuen,
2003). Since the viscosity of the surrounding dry mantle controls the propagation of
the blob, a certain mechanism is responsible for the occurrence of the low
wrapping viscosity.
As the blob ascents through the mantle wedge, porous flow (Davies and
Stevenson, 1992) of the melt and fluid might penetrate the mantle around the blob,
- 74 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
dropping there the viscosity down to 1014 Pa s. The rate of this penetration by
porous flow has to be at least the same as the ascending rate of the blob. Recent
laboratory experiments by Hall and Kincaid (2001) show that the positively buoyant
blobs might pass through the low-viscosity, low-density paths created by previous
blobs. The mechanism that produces the low wrapping viscosities may be a
combination of the porous flow and a low viscosity conduit left by previous blobs.
Since the rigid sphere approximation is used to estimate the drag force
applied to weak partially molten blobs, the estimates of wrapping viscosity in the
present study should be considered as rather approximate. The blobs of 10 km
diameter reach the base of the continental crust in ~ 6 Myr when the wrapping
viscosity is of ~ 3·1017 Pa s. Since this viscosity value is in the lower range
reported for dry mantle (Moore et al., 1998), no wrapping viscosity mechanism is
needed in this case. The blobs of such big size can affect locally the mantle wedge
flow pattern and the effective viscosity distribution (Gerya and Yuen, 2003),
especially for the temperature and strain-rate dependent viscosity (Ranalli, 1995).
These effects are not included in the present models.
The time required for the blob of 1 km diameter to rise from the slab surface
up to the continental crust varies between 0.001 Myr and 14 Myr for the viscosity
between 1014 Pa s and 5·1017 Pa s respectively. In general, the blob rising time
decreases nonlinearly as its diameter increases and the wrapping viscosity is
diminishing (Fig. 16). The reference viscosity, η0, in the range up to 1020 Pa s, has
a negligible effect on the blob trajectory. The effect can be noticed for the higher
values of the reference viscosity (1021 Pa s) when the raising time is more than ~ 2
Myr (Fig. 16).
The blob tracing dynamic model in the mantle wedge velocity field shows
that the “fast” trajectories end at the same focus location (below the Popocatépetl
volcano, ~ 350 km from trench) on the base of the continental crust (Fig. 15). This
result may be interpreted as a possible condition for the development of the
stratovolcanoes. The ending points of “slow” trajectories, which are common for the
blobs of smaller size (~ 0.4 - 0.6 km), are scattered from the focus location (Fig.
- 75 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
15-A). This observation may give us a hint on a possible mechanism of
monovogenic volcanism.
Recent studies (Lucia Capra – 2004, personal communication) revealed at
least two magmatic pulses with a time span of ~ 1 Myr on the stratovolcano
Nevado de Toluca. The maximum volume of each of these magmatic events is of
3.5 km3 (equivalent blob diameter is ~ 2 km). It is interesting that both pulses
started with a magmatic signature of melted sediments (Ce anomaly). These new
studies will provide an important constrains on the blob size, rising time and the
blob composition. From our model (Fig. 15), a blob of 2 km in diameter reaches the
base of the continental crust in 1 Myr, if the wrapping viscosity is of ~ 2·1015 Pa s.
The low viscosity is essential for the smaller size blobs to rise to the base of the
continental crust.
The average volume of a monogenic cinder cone in the CMVB is less than 1
3
km (Hasenaka, 1994), which corresponds to blob diameters of ~ 1.3 km. If the
origin of monogenic cones are described by the blob tracing model, then the
wrapping viscosity should be of η > 5·1015 Pa s to produce the “slow” trajectories.
We need further model enhancement to verify the relation between the blob (or
plume) hypothesis and the origin of the strato and monogenic volcanism in the
CMVB.
- 76 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
REFERENCES
Alaniz-Alvarez, S.A., Nieto-Santiago, A.F., Tolson, G., 1998. A graphical
technique to predict slip along a preexisting plane of weakness. Engineering
Geology, v. 49, pp.53-60.
Arzate, J.A., Mareschal, M. and Urrutia-Fucugauchi, J., 1993. A preliminary
crustal model of the Oaxaca continental margin and subduction zone from
magnetotelluric and gravity measurements. Geofísica Internacional, 32, 441-45.
Blatter, D.L., and Carmichael, I.S.E., 1998. Hornblende peridotite xenoliths from
central Mexico reveal the highly oxidized nature of subarc upper mantle. Geology,
v. 26, no. 11, pp. 1035-1038.
Burov, E., Jolivet, E., Le Pourhiet, L. and Poliakov, A., 2000. A
thermomechanical model of exhumation of high pressure (HP) and ultra-high
pressure (UHP) metamorphic rocks in Alpine-type collision belts. Tectonophysics,
342: 113-136.
Byerlee, J.D., 1978. Friction of rocks. Pure Applied Geophysics, 116, 615-626.
Cebull, S.E., and Schubert, D.H., 1987. Mexican Volcanic Belt-an intraplate
transform. Geofisica Internacional, v.26, pp. 1-13.
Conder, J.A., Weins, D.A., and Morris, J., 2002. On the decompression melting
structure at volcanic arcs and back-arc spreading centres. Geophysical Research
Letters,29, 17-1-17-4.
Currie C.A., Hyndman, R.D., Wang, K. and Kostoglodov, V., 2002. Thermal
models of the Mexico subduction zone: Implications for the megathrust
seismogenic zone. Journal of Geophysical Research, 107, NO. B12, 2370,
doi:10.1029/2001JB000886
Davies, J.H., and Stevenson,D.J., 1992. Physical model of source region of
subduction zone volcanism. Journal of Geophysical Research, 97, 2037-2070.
DeMets, C., Gordon, R., Argus, D. and Stein, S., 1994. Effect of recent revisions
to the geomagnetic reversal time scale on estimates of current plate motions.
Geophysical Research Letters, 21, 2191-2194.
- 77 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Fedotov, S.A., 1981. Magma rates in feeding conduits of different volcanic
centres. Journal of Volcanology and Geothermal Research, v. 9, pp. 379-394.
Ferrari, L., Garduno, V., Innocenti, F., Manetti, P., Pasquere, G., and Vaggeli,
G., 1994. Volcanic evolution of central Mexico: Oligocene to Present. Geofisica
Internacional, 33, 91-105.
Ferrari, L., Pasquare, G., and Tybaldi, A., 1990. Plio-Quaternary tectonics on the
central Mexican Volcanic Belt and some constraints on its rifting mode. Geofisica
Internacional, v. 29, pp. 5-18.
Furukawa, F., 1993. Magmatic processes under arcs and formation of the volcanic
front. Journal of Geophysical Research, 98, 8309-8319.
Gerya, T.V. and Yuen, D.A., 2003. Rayleigh-Taylor instabilities from hydration and
melting propel ‘cold plumes’ at subduction zones. Earth and Planetary Science
Letters, 212, 47-62.
Gomez-Tuena A., LaGatta, A.B., Langmuir, C.H., Gutierrez, F.O., CarrascoNunez, G., 2003. Temporal control of subduction magmatism in the eastern TransMexican
Volcanic
Belt:
Mantle
sources,
slab
contributions,
and
crustal
contamination. G-cubed, vol.4, no. 8.
Gorbatov A., and V. Kostoglodov, 1997. Maximum depth of seismicity and thermal
parameter of the subducting slab: worldwide empirical relation and its application.
Tectonophysics, 227, 165-187
Hacker, B.R., Abers, G.A., and Peacock, S.M., 2003, Subduction Factory 1.
Theoretical mineralogy, densities, seismic wave speeds, and H2O contents.
Journal of Geophysical Research, v. 108, 10.1029/2001JB001127.
Hall, P., and Kincaid, C., 2001. Diapiric flow at subduction zones: a recipe for
rapid transport. Science, 292, 2472-2475.
Hasenaka, B.R., 1994. Size, distribution, and magma output rate for shield
volcanoes of the Michoacan-Guanajuato volcanic field, central México. Journal of
Volcanalogy and Geothermal Res., 63, 13-31.
- 78 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Hirth, G., and Kohlstedt, D.L., 1995. Experimental constraints on the dynamics of
the partially molten upper mantle: Deformation in the diffusion creep regime.
Journal of Geophysical Research, 100, 1981-2001.
Hirth, G., and Kohlstedt, D.L., 2003. Rheology of the mantle wedge: in Inside the
Subduction Factory, Geophysical Monograph 138, edited by J. Eiler, AGU,
Washington DC, p. 83-105.
Karato, S., and Wu, P., 1993. Rheology of the upper mantle: a synthesis. Science,
260, 771-778.
Kelemen, P.B., Rilling, J.L., Parmentier, E.M., Mehl, L., and Hacker, B.R.,2003.
Thermal Structure due to Solid-State Flow in the Mantle Wedge Beneath Arcs, in
Inside the Subduction Factory, Geophysical Monograph 138, edited by J. Eiler,
AGU, Washington DC, pp. 293-311.
Klein, E.M., and Langmuir, C.H., 1987. Global correlations of ocean ridge basalt
chemistry with axial depth and crustal thickness. Journal of Geophysical Research,
92, 8089-8115.
Klitgord, K. and Mammerickx, J., 1982. Northern East Pacific Rise: Magnetic
anomaly and bathymetric framework. Journal of Geophysical Research, 87, 67256750.
Kostoglodov V. and W. Bandy, 1995. Seismotectonic constraints on the
convergence rate between the Rivera and North American plates. Journal of
Geophysical Research, 100, 17,977-17,989.
Kostoglodov V., W. Bandy, J. Domínguez, and M. Mena, 1996. Gravity and
seismicity over the Guerrero seismic gap, México. Geophysical Research Letters, 23,
3385-3388.
Kostoglodov, V., Singh, S. K., Santiago, J. A., Franco, S. I., Larson, K. M.,
Lowry, A. R., and Bilham, R., 2003. A large silent earthquake in the Guerrero
seismic
gap,
Mexico.
Geophysical
Research
Letters,
30(15),
1807,
doi:10.1029/2003GL017219.
Langmuir, C.L., Klein, E.M., and Plank, T., 1992. Petrological systematics of midocean ridge basalts: Constraints on melt generation beneath ocean ridges, in
- 79 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Mantle flow and melt generation at mid-ocean ridges. AGU Monograph 71, edited
by J. Phipps Morgan, D.K. Blackman, and J.M. Sinton, pp. 183-280, AGU,
Washington DC.
Luhr, J.F., 1997. Extensional tectonics and the diverse primitive volcanic rocks in
the western Mexican Volcanic Belt. Canadian Mineralogy, 35: 473-500.
Manea, M., Manea, V.C., Kostoglodov, V., 2003. Sediment Fill of the Middle
America Trench Inferred from the Gravity Anomalies. Geofisica Internacional, 42,
(4), 603-612.
Manea, V.C., Manea, M., Kostoglodov, Currie, C.A., and Sewell, G., 2004.
Thermal Structure, Coupling and Metamorphism in the Mexican Subduction Zone
beneath Guerrero. Geophysical Journal International, 158, 775–784 doi:
10.1111/j.1365 - 246X.2004.02325.x.
Márquez, A., Oyarzun, R., Doblas, M., Verma, S.P., 1999a. Alkalic (OIB-type)
and calc-alkalic volcanism in Mexican volcanic belt: a case study of plume-related
magmatism and propagating rifting at an active margine? Geology, 27: 51-54.
Márquez, A., Oyarzun, R., Doblas, M., Verma, S.P., 1999b. Replay to comment
to: Alkalic (OIB-type) and calc-alkalic volcanism in Mexican volcanic belt: a case
study of plume-related magmatism and propagating rifting at an active margine?
Geology, 27: 1055-1056.
Márquez, A., Verma, S.P., Anguita, F., Oyarzun, R., Brandle, J.L., 1999c.
Tectonics and volcanism of Sierra Chichinautzin: extension at the front of the
central Trans-Mexican Volcanic Belt. Journal of Volcanology and Geothermal
Research, 93, 125-150.
Márquez, A. and De Ignatio, C., 2002. Mineralogical and geochemical constrains
for the origin and evolution of magmas in Sierra Chichinautzin, Central Mexican
Volcanic Belt. Lithos, 62, 35-62.
Moore, J.C., J.S. Watkins, T.H.Shipley, K.J. McMillen, S.B. Bachman, and N.
Lundberg, 1982. Geology and tectonic evolution of a juvenile accretionary terrane
along a truncated convergent margin: Synthesis of results from Leg 66 of the Deep
- 80 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Sea Drilling Project, southern México. Geological Society American Bulletin, 93,
847-861.
Moore, W. B., Schubert, G., and Tackley, P. J., 1998. Three-Dimensional
Simulations of Plume-Lithosphere Interaction at the Hawaiian Swell. Science, 279,
1008-1011.
Nichols, G.T., Wyllie, P.J., and Stern, C.R., 1994. Subduction zone melting of
pelagic sediments constrained by melting experiments. Nature, 371, 785-788.
Pardo, M. Suárez, G., 1995. Shape of the subducted Rivera and Cocos plates in
southern mexico: seismic and tectonic implications. Journal of Geophysical
Research, 100, 12357-12373.
Peacock, S.M., and K. Wang, 1999. Sesimic consequences of warm versus cool
subduction metamorphism: Examples from southwest and northeast Japan.
Science, 286, 937-939.
Prol-Ledesma, R.M., Sugrobov, V.M., Flores, E.L., Juarez, G., Smirnov, Y.B.,
Gorshkov, A.P., Bondarenko, V.G., Rashidov, V.A., Nedopekin, L.N. and
Gavrilov, V.A., 1989. Heat flow variations along the Middle America Trench.
Marine Geophysical Research, 11, 69-76..
Ranalli, G., 1995. Rheology of the Earth, 2nd edition, Chapman and Hall, London,
413pp.
Shubert, D.H., and Cebull, S.E., 1984. Tectonic interpretation of the TransMexican volcanic belt. Tectonophysics, 101, pp.159-165.
Smith, D.L., Nuckels, C.E., Jones, R.L. and Cook, G.A., 1979. Distribution of
heat flow and radioactive heat generation in northern Mexico, Journal of
Geophysical Research, 84, 2371-2379.
Takada, A., 1989. Magma transport and reservoir formation by a system of
propagating cracks. Bulletin of Volcanology, v. 52, pp.118-126.
Takada, A., 1994. The influence of regional stress and magmatic input on styles of
monogenetic and polygenetic volcanism. Journal of Geophysical Research, v. 99,
pp. 13563-13573.
- 81 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Tatsumi, Y., 1986. Formation of the volcanic front in subduction zones.
Geophysical Research Letters, 13, 717-720.
Turcotte, D.L., and Schubert, G., 1982. Geodynamics, Applications of continuum
Physics to Geological Problems. Wiley, New York, NY.
Vacquier, V., J.G. Sclater, and C.E. Corry, 1967. Studies of the thermal state of
Earth. The 21st paper: Heat-flow, eastern Pacific. Bulletin of Earthquake Research
Institute, 45, 375-393.
Valdes, C.M., Mooney, W.D., Singh, S.K., Meyer, R.P., Lomnitz, C, Luetgert,
J.H., Helsley, C.E., Lewis, B.T.R. and Mena, M., 1986. Crustal structure of
Oaxaca, Mexico, form seismic refraction measurements. Bulletin of Seismological
Society of America, 76, 547-563.
van Hunen, J., van den Berg, A., and Vlaar, N., 2002. On the role of subducting
oceanic plateaus in the development of shallow flat subduction. Tectonophysics,
352, 317-333.
van Keken, P.E., Kiefer, B., and Peacock, S.M., 2002. High resolution models of
subduction zones: Implications for mineral dehydration reactions and the transport
of water into deep mantle. G-cubed, 3, 10, 20.
Verma, S.P., 1999. Geochemistry of evolved magmas and their relationship to
subduction un-related mafic volcanism at the volcanic front of the central Mexican
Volcanic Belt. Journal of Volcanology and Geothermal Research, v.93, pp. 151171.
Verma, S.P., 2000. Geochemistry of subducting Cocos plate and the origin of
subduction-unrelated mafic volcanism at volcanic front of central Mexican Volcanic
Belt, in Delgado-Granados, H., Aquirre-Diaz, G.J., and Stock, J., eds.,
Cenozoictectonics and volcanism of Mexico: Boulder Colorado, Geological Society
of America, Special Paper, 334, pp. 195-222.
Wallace, P., Carmichael, I.S.E., 1999. Quaternary volcanism near the Valley of
Mexico: implications for subduction zone magmatism and the effects of crustal
thickness variations on primitive magma compositions. Contribution to Mineral
Petrology. 135, 291-314.
- 82 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Wang, K. and Davis, E.E., 1992. Thermal effect of marine sedimentation in
hydrothermally active areas. Geophysical Journal International, 110, 70-78.
Wyllie, P.J., 1979. Magmas and volatile components, American Mineralogy, 654,
469-500.
Ziagos, J.P., D.D. Blackwell, and F. Mooser, 1985. Heat flow in southern Mexico
and the thermal effects of subduction, Journal of Geophysical Research, 90, 54105420.
- 83 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Table 1. Summary of the thermal parameters used in the models. (Compilation
from: Peacock and Wang, 1999; Smith et al., 1979; Ziagos et al., 1985; Vacquier et
al., 1967; Prol-Ledesma et al., 1989).
Geological Unit
Density
(kg/m3)
Thermal
Conductivity
(W/mK)
Radiogenic
heat
production
3
(µW/m )
Thermal
Capacity
(MJ/m3K)
2200
1.00 – 2.00*
1.00
2.50
2700
2.00
0.65
2.50
2700
2.00
0.15
2.50
Mantle wedge
3100
3.10
0.01
3.30
Oceanic lithosphere
3000
2.90
0.02
3.30
Oceanic sediments
Upper continental crust
(0-15 Km)
Lower continental crust
(15-40 Km)
* Increase linearly with distance from the deformation front up to a depth of
10 Km.
- 84 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
FIGURE CAPTIONS
Figure 1.
Tectonic setting and position of the modeled cross-section (straight line), in
Guerrero. Triangles show the location of active volcanoes in Mexico. Dashed
ellipse is the CMVB - Central Mexican Volcanic Belt. Light blue rectangle
represents the El Peñón area, where mantle xenoliths have been found. Arrows
show convergence velocities between the Cocos and North American plates
(DeMets et al., 1994).
Figure 2.
The grid with 12,000 elements used to solve the numerical models. This
particular shape of the grid is designed to better resolve the velocity field nearby
the tip of the mantle wedge.
Figure 3.
Boundary conditions and parameters used in the modeling. The upper and
lower boundaries have constant temperatures of 0ºC and 1,450ºC, accordingly.
The continental plate is fixed. The right (landward) vertical boundary: 20ºC/km
thermal gradient in the continental crust (down to 40 km); between 40 km and 100
km depth the thermal gradient is of 10ºC/km; no horizontal conductive heat flow is
specified beneath 100 km depth. Zero traction is considered beneath Moho (40
Km), at the boundary, which belongs to the mantle wedge. The convergence
velocity is specified for the intersection between the right boundary and the
subducting slab. The left (seaward) boundary condition is a one-dimensional
geotherm for the oceanic plate. The Cocos plate motion is referred to the North
American plate with the convergence velocity of 5.5 cm/yr. Volumetric shear
heating is imposed along the plate interface up to a maximum depth of 40 km,
using the Byerlee's friction law (Byerlee, 1978).
- 85 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4.
Calculated steady-state thermal field for the isoviscous mantle wedge.
Horizontal black dashed line shows the Moho (40 km depth). The hinge point is at
270 km from the trench. Thick violet solid line denotes the top of the subducting
slab. Orange triangles are the two principal stratovolcanoes in CMVB: Nevado de
Toluca and Popocatépetl. The maximum temperature beneath Popocatépetl
volcano is about of 830°C. The lower left inset is the magnified thermal structure
close to the tip of the mantle wedge. Thin dark blue arrows in the mantle wedge
represent the velocity field. The intraslab earthquakes with magnitudes, Mw ≥ 5.5,
are represented by the focal mechanisms. The two brown clouds of hypocenters
beneath the coast denote the smaller magnitude seismicity associated probably
with the bending of the subducted plate.
Figure 5.
Variation of the surface heat flow along the Guerrero profile. The dots with
vertical error bars are heat flow data reported by Ziagos et al. (1985).
(A). Surface heat flow for steady-state thermal models reference viscosities
η0 = 1017 - 1021 Pa s and Ea = 300 kJ/mol.
(B). Surface heat flow for steady-state thermal models activation energies
Ea = 150 - 350 kJ/mol and η0 = 1020 Pa s.
Figure 6.
Steady-state thermal models with strong temperature-dependent viscosity in
the mantle wedge. The reference viscosity, η0, varies from 1017 Pa s (A) up to 1021
Pa s (E). The activation energy for olivine of 300 kJ/mol is used. Note an important
increase of the temperature close to the tip of the wedge (A-D). NP and P
represent the Nevado de Toluca and Popocatépetl volcanoes. The maximum
temperature beneath the Popocatépetl volcano is about of 1,230°C. Other
notations are the same as in Fig. 4.
- 86 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 7.
Steady-state thermal models with strong temperature-dependent viscosity in
the mantle wedge for activation energies, Ea, from 150 kJ/mol (A.) to 350 kJ/mol
(E). The reference viscosity of 1020 Pa s is used. The maximum temperature
beneath the Popocatépetl volcano is about of ~ 1,250°C (A - D) and more than 1
300°C (E). Other notations are the same as in Fig. 4.
Figure 8.
(A). Phase diagrams for the MORB (Hacker et al., 2002). 1 - Zeolite (4.6
wt% H2O), 2 - Prehnite - Pumpellyite (4.5 wt% H2O), 3 - Pumpellyite - Actinolite
(4.4 wt% H2O), 4 - Greenschist (3.3 wt% H2O), 5 - Lawsonite - Blueschist (5.4 wt%
H2O), 6 - Epidote - Blueschist (3.1 wt% H2O), 7 - Epidote - Amphibolite (2.1 wt%
H2O), 8 - Jadeite - Epidote - Blueschist (3.1 wt% H2O), 9 - Eclogite - Amphibole
(2.4 wt% H2O), 10 - Amphibolite (1.3 wt% H2O), 11 - Garnet - Amphibolite (1.2 wt%
H2O), 12 - Granulite (0.5 wt% H2O), 13 - Garnet - Granulite (0.0 wt% H2O), 14 Jaedite - Lawsonite - Blueschist (5.4 wt% H2O), 15 - Lawsonite - Amphibole Eclogite (3.0 wt% H2O), 16 - Jaedite - Lawsonite - Talc - Schist, 17 - Zoisite Amphibole - Eclogite (0.7 wt% H2O), 18 - Amphibole-Eclogite (0.6 wt% H2O), 19 Zoisite-Eclogite (0.3 wt% H2O), 20 - Eclogite (0.1 wt% H2O), 21 - Coesite - Eclogite
(0.1 wt% H2O), 22 - Diamond - Eclogite (0.1 wt% H2O). Slab surface geotherms
are calculated for the reference viscosity range: 1017 - 1021 Pa s (see inset).
(B). Phase diagram for harzburgite (Hacker et al., 2002). A - Serpentine Chlorite - Brucite (14.6 wt% H2O), B - Serpentine - Chlorite - Phase A (12 wt%
H2O), C - Serpentine - Chlorite - Dunite (6.2 wt% H2O), D - Chlorite - Harzburgite
(1.4 wt% H2O), E - Talc - Chlorite - Dunite (1.7 wt% H2O), F - Anthigorite - Chlorite
- Dunite (1.7 wt% H2O), G - Spinel - Harzburgite (0.0 wt% H2O), H - Garnet Harzburgite (0.0 wt% H2O). Calculated geotherms for the slab surface are the
same as in A. The vertical temperature profile beneath the Popocatépetl volcano
are the same as in Fig. 10. The amount of serpentine (green-yellow hatched
insets) in the tip of the mantle wedge is noticeably smaller for the model with
- 87 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
temperature dependent viscosity (upper left inset) than for the model with the
isoviscous mantle wedge (upper right inset).
Figure 9.
The results of benchmark test with different grid resolutions: 4,000, 6,000,
8,000, 1,000 and 12,000 triangles. The temperature differences between two
subsequent models are calculated along the slab-wedge interface. Note the
temperature fluctuation up to 22°C for low grid resolution (blue curve). Less than
5°C numerical error results when the grid with 12,000 triangles is used (red curve).
Figure 10.
Vertical temperature profile (A-A’) for the mantle wedge thermal models with
the reference viscosities of 1017 - 1021 Pa s. The activation energy is fixed at 300
kJ/mol. The temperature profile for the model with the isoviscous mantle is also
shown as a dashed line.
Figure 11.
Distribution of the mantle wedge viscosity for steady-state thermal (Fig. 4)
models with strong temperature-dependent viscosity. The reference viscosity, η0, is
from 1017 Pa s (A) up to 1021 Pa s (E).
Figure 12.
Vertical temperature profile (A-A’) through the mantle wedge. The thermal
modeling is done for activation energies of 150 - 350 kJ/mol. The reference
viscosity is fixed at 1020 Pa s. Dashed line shows, for a comparison, the
temperature profile for the model with the isoviscous mantle wedge.
- 88 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 13.
Distribution of mantle wedge viscosity with strong temperature-dependence
for steady-state thermal models (Fig. 12). Activation energies, Ea, are from 150
kJ/mol up to 350 kJ/mol.
Figure 14.
(A). Phase diagrams for the MORB and maximum H2O contents (Hacker et
al., 2002). Modeled slab surface geotherms are plotted for the activation energy
range: 150 - 350 kJ/mol (see inset). The other notations are the same as in Fig. 8.
(B). Phase diagram for harzburgite, and maximum H2O contents (Hacker et
al., 2002). The calculated geotherms are the same as in A, as well as those from
the vertical profile beneath Popocatépetl (see Fig. 13). The amount of serpentine
(green-yellow hatch upper insets) in the tip of the wedge decreases as the
activation energy increases.
Figure 15. Blob trajectories in the steady mantle wedge flow Initial points for all
trajectories are selected on the surface of the slab, right below the Popocatépetl
volcano at 70 km depth.
(A). The wrapping viscosity is fixed at 1015 Pa s. The blobs with the diameter
less than 0.4 would never rise up to the continental crust.
(B). The blob diameter is fixed at 10 km. The trajectories corresponding to
different wrapping viscosities of less than 1017 Pa s have the same final point below
the Popocatépetl volcano. The blobs would never rise up to the continental crust if
with the wrapping viscosities is > 5·1017 Pa s.
(C). The blob’s diameter is fixed at 2 km. The trajectories corresponding to
different wrapping viscosities of less than 9·1015 Pa s, have the same final point
below the Popocatépetl volcano. The blobs with the wrapping viscosities > 2·1016
Pa s would never rise up to the continental crust.
(D). The wrapping viscosity is fixed at 1017 Pa s. The trajectories correspond
to different blob size (4 - 10 km). The blobs with the diameter less than 4 km would
never rise up to the continental crust.
- 89 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 16.
Blob rising time as a function of the wrapping viscosity. For a wide range of
rheological parameters (η0 = 1017 - 1020 Pa s and Ea = 150 - 350 kJ/mol) the blob
trajectories and rising times are nearly identical (blue curves). For the reference
viscosity of 1021 Pa s, the rising time becomes longer after ~ 1 Myr, for a given
wrapping viscosity (dashed green curves). The curves annotated with the blob’s
diameter show that the rising time is decreasing for the bigger blobs and the lower
viscosity. The blobs with the size of ~ 2 km can reach the continental crust in less
than 1 Myr (black dashed line) at a wide range of low wrapping viscosities.
Figure 17.
(A). Calculated slab surface geotherms for models with reference viscosities
in the range of 1017 - 1021 Pa s (see inset), and fluid saturated sediment solidus
from (Nichols et al., 1994). The horizontal dashed arrows mark the depths where
the sediment meting might occur. Dashed line represents the geotherm for the
isoviscous mantle.
(B). Calculated slab surface geotherms for models with activation energy
range: 150 - 350 kJ/mol (see inset) and fluid saturated sediment solidus from
Nichols et al., 1994. The horizontal dashed arrows mark the depths where
sediment meting might occur.
- 90 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 1
- 91 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 2
- 92 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 3
- 93 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4
- 94 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 5
- 95 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 6
- 96 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 7
- 97 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 8
- 98 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 9
- 99 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 10
- 100 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 11
- 101 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 12
- 102 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 13
- 103 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 14
- 104 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 15
- 105 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 16
- 106 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 17
- 107 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
IV. MODELOS TÉRMICOS, TRANSPORTE DEL MAGMA Y ESTIMACIÓN DE LA
ANOMALÍA DE VELOCIDAD DEBAJO DE KAMCHATKA MERIDIONAL
Accepted in “Melting Anolmalies: Their Nature and Origin”,
GSA Post-Conference Book
THERMAL MODELS, MAGMA TRANSPORT AND VELOCITY ANOMALY
ESTIMATION BENEATH SOUTHERN KAMCHATKA
V.C. Manea1, M. Manea1, V. Kostoglodov1, G. Sewell2
1
Instituto de Geofisica, Universidad Nacional Autónoma de México (UNAM),
México
2
University of Texas, El Paso
ABSTRACT
A finite element method is applied to model the thermal structure of the
subducted Pacific plate and overlying mantle wedge beneath the southern part of
the Kamchatka peninsula. A numerical scheme solves a system of 2D NavierStokes equations and a 2D steady state heat transfer equation.
A model with isoviscous mantle exposed very low temperatures (~ 800ºC) in
the mantle wedge, which cannot account for magma generation below the volcanic
belt. Instead, a model with strong temperature-dependent viscosity shows a rise in
the temperature in the wedge. At a temperature of more than 1300ºC beneath the
active volcanic chain, melting of wedge peridotite becomes possible. Although the
subducting slab below the Kamchatka peninsula is rather old (~ 70 Myr), some
frictional heating (µ = 0.034) along the interface between the subducting oceanic
slab and the overlying Kamchatka peninsula lithosphere would be enough to melt
subducted sediments. Dehydration (> 5 wt% H2O release) occurs in the subducting
- 108 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
slab because of metamorphic changes. As a consequence, hydration of the mantle
wedge peridotite might produce melt, which may rise to the base of the continental
crust as diapir-like blobs.
Considering that melting processes in the subducting plate generate most of
the volcanic material, we developed a dynamic model which simulates the
migration of partially melted buoyant material in the form of blobs in the viscous
mantle wedge flow. Blobs with a diameter of 0.4 - 10.0 km rise to the base of the
continental lithosphere within 0.002 - 10 Myr depending on the blob diameter and
surrounding viscosity.
The thermal structure obtained in the model with temperature dependent
viscosity is used to estimate seismic P-wave velocity anomalies (referenced to
PREM) associated with subduction beneath Kamchatka. A low velocity zone (~ 7% velocity anomaly) is obtained beneath the volcanic belt and a high velocity
anomaly (~ 4%) for the cold subducted lithosphere. These results agree with
seismic tomography results from P-wave arrivals.
Keywords: Kamchatka subduction zone, thermal models, mantle wedge flow,
blobs, tomographic imaging.
- 109 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
GENERAL ASPECTS OF SUBDUCTION ZONE DYNAMICS
Mantle plumes are frequently assumed to come from the transition zone or
from the core-mantle boundary. There is another place inside the Earth from where
these thermal or chemical instabilities might develop: the top layer of the
subducting slabs where strong dehydration and melting is expected. This chapter
aims to study the subduction magmatism through numerical modeling of the mantle
wedge thermal structure and magma transport with a diapir model.
The thermal structure of subduction zones provides important insight into
the thermal and chemical exchange between subducted oceanic lithosphere and
overlying mantle wedge, and on magma generation and transport. Various thermal
models have been developed based on analytical approximations (Davies, 1999;
Molnar and England, 1995, 1990), analytical solutions for wedge corner flow
(Peacock, 2002, 1996, 1991, 1990a, 1990b; Peacock and Hyndman, 1999;
Peacock and Wang, 1999; Billen and Gurnis, 2001), and recent models which
incorporate temperature - and stress - dependence viscosity (von Hunen, 2000,
van Keken, 2002, Gerya and Yuen, 2003).
The object of the present study is to model one of the oldest and most
interesting subduction zones; the Kamchatka subduction zone (KSZ), in which
thermal structure is poorly known up to now. Kamchatka thermal structure is
analyzed in models with constant viscosity (isoviscosity) and strong temperaturedependent viscosity of the asthenosphere.
A distinctive feature of the thermal regime associated with subduction zones
is the inverted thermal gradient just above the slab in the mantle wedge where the
most important and intensive chemical and thermal exchanges occur. In this region
the material transferred from the oceanic slab to the mantle wedge enriches the
source region of arc lavas with incompatible elements and volatiles (Plank and
Langmuir, 1993; Stolper and Newman, 1994; Johnson and Plank, 1999; Eiler et al.,
2000). An influx of volatiles from metamorphosed oceanic crust and sediments
triggers partial melting of peridotite above the subducted slab (Tatsumi, 1986;
- 110 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Davies and Stevenson, 1992). Basically, three mechanisms of melt generation are
expected: partial melting producing positively
buoyant diapirs (“blobs” in the
present paper) (Tatsumi et al., 1983; Kushiro, 1990; Gerya and Yuen, 2003;
Manea et al., 2004b); anhydrous decompression melting of peridotite (Klein and
Langmuir, 1987; Langmuir et al., 1992); porous flow of hydrated partial melt
(Gaetani and Grove, 2003).
The latest insights revealed by new developed high-resolution dynamical
models for subduction zones with 0.5 billion markers of Gerya and Yuen (2003)
show that the hydration and related melting introduce a source of chemical
buoyancy on the 1 - 10 km thick layer on top of the subducting slab. These
plumes, with diameter up to 10 km, penetrate the hot mantle wedge with velocities
of ~ 10 cm/yr.
Various mechanisms were proposed for rapid magma transport from the
mantle wedge toward the surface: through channelized network (channels
diameter of 1 - 100 m) (Spiegelman and Kelemen, 2003), through fractures and
shear zones in the mantle (Shaw, 1980), through a fractal tree network (very thin
channels with 1 - 4 mm diameter) (Hart, 1993) and the dike transport mechanism in
the mantle (Rubin, 1993). The model of Shaw (1980) proposes fracture zones
inside the mantle wedge. The presence of such fracture system is unlikely to exist
in a very hot mantle wedge (> 1300°C). The mantle wedge peridotite behaves
completely ductile for temperature above ~ 700°C. Spiegelman and Kelemen
(2003) provide a model capable to explain the chemical and spatial variability of
lava samples taken from mid ocean ridges. The geodynamics of subduction zones
is quite different from the one of mid ocean ridges, basically the strong convection
beneath the volcanic arcs represents one of the main differences. The applicability
of this model with magma transport velocity of ~ 1.5 m/year through channels with
diameters < 100 m proposed by Spiegelman and Kelemen (2003) needs to be
further investigated in the presence of strong convecting systems as the mantle
wedge beneath volcanic arcs. Although the other models (Hart, 1993; Rubin, 1993)
predict fast magma transport toward the surface through very narrow channels,
- 111 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
they do not take into account the presence of strong mantle wedge convection
which affect the continuation of these channels. For this reason, in the present
study, for magma generation and propagation in the southern Kamchatka
subduction zone, concept of positive buoyant diapirs (blobs) that migrate in the
mantle wedge flow induced by the subducting slab is applied. We used for the
modeling the numerical scheme of Manea et al. (2004b).
Partial melting of fluid-saturated peridotite at low temperature on top of the
subducted slab may generate H2O-rich partial melt. In order for melt to progress,
ascending blobs have to be efficiently heated as they move in the mantle wedge
(Gaetani and Grove, 2003). Conductive thermal evolution of these diapirs as they
ascend toward the base of the continental crust is modeled.
Mantle wedge viscosity is an important factor that controls the migration of
diapirs in the mantle wedge. Dissolved H2O and the partial melt might lower mantle
viscosity below 1018 Pa s (Kelemen et al., 2003). Moreover, Hirth and Kohlstedt
(2003) showed that experimentally measured viscosities for olivine at upper mantle
pressures and temperatures are in the range 1017 - 1021 Pa s.
Thermal modeling is applied to investigate the possibility of melting slab fluidsaturated sediments (wet solidus from Nichols et al., 1994) and mantle-wedge wet
peridotite (wet solidus from: Mysen and Boettcher, 1975; Wyllie, 1979), and
dehydration along the slab-wedge interface beneath Kamchatka.
The main constraint on mantle-wedge temperature distribution could be Pwave seismic tomography. Zhao et al. (1992, 1997) observed a low-velocity zone (6% P-wave velocity anomaly) beneath the Tonga and NE Japan volcanic arcs,
which might be related to upper-mantle melting below the base of the continental
crust. Gorbatov et al. (1999) detected a low-velocity zone (-7% P-wave velocity
perturbation) beneath the active volcanic chain in Kamchatka. Tomography images
of Northeast Japan (Tamura et al., 2002; Zhao, 2001) reveal a limited low velocity
region that is in contact directly with the slab surface (-2 - 3% velocity anomaly) at
depths of 80 - 100 km.
- 112 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Following Karato (1993), the temperature dependence of seismic wave
velocities can be used to estimate velocity anomalies from thermal models. An
agreement between the velocity perturbation beneath the volcanic arc (shape and
magnitude) observed in the seismic tomography anomaly and the tomography
anomaly estimation from thermal modeling might be an indication of a satisfactory
estimate of the thermal regime in the mantle wedge.
- 113 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
REGIONAL TECTONIC SETTINGS
The Kamchatka peninsula has one of the most active volcanic chains in the
world and is a dynamic convergent margin where the Pacific plate (PAC) subducts
beneath the North-American plate. The PAC subducts westward at a dip angle of ~
55º from 50ºN to 54ºN, and at a rate of ~ 7.8 cm/year (Gorbatov et al., 1999). The
age of the subducting plate along the Kamchatka trench varies between 70 and
100 Myr. The dip of the Wadati-Benioff zone varies along the KSZ from ~ 55º in the
south to ~ 35º in the north. Heat flow data (Smirnov and Sugrobov, 1979; 1989a,b)
suggest that the thermal thickness of the subducted plate is up to 75% smaller in
the north of the peninsula then in the south. Thus the effective age (thermally
defined) is less there, because of the geological age (Renkin and Sclater, 1988). In
order to avoid the effect of oceanic plate rejuvenation in the north of the KSZ, the
present paper explores a 2D profile normal to the trench in the southern part of
Kamchatka (Fig. 1), far from this important thermal anomaly. The seismicity and
structure of the KZS was studied in detail by Gorbatov et al. (1997). The volcanic
front, from 50ºN to 54ºN, is trench-parallel and corresponds to a depth of the
subducted slab of about 90 to 140 km.
Southern Kamchatka is separated into en eastern active volcanic chain and
the western inactive tectonically and volcanically Sredinny Range (Fedotov &
Masurenkov, 1991). Numerous active and inactive volcanoes, which form the
Eastern Volcanic Front (EFV), are situated above the subducting slab at depth
where partial melting in the mantle wedge is induced by fluids from slab
dehydration.
Observations of a slab-melt chemical signature are mainly restricted to the
northern volcanoes as Sheveluch, which is the only active volcano with an adakite
magma signature. The tectonic reconstruction of Kamchatka-Aleutian corner (Park
et al., 2002) strongly suggests that the extinct subduction zone (just beyond the
edge of the slab near the Aleutian junction) involved a shallow-dipping young slab.
The melting of subducted slab and fluids from slab dehydration produced an
- 114 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
adakites type of arc volcanism. However, in southern Kamchatka no adakites were
found, suggesting that in this region the slab is not melting. The lack of this style of
volcanism would be a very important constraint on the subducting slab thermal
structure. While melting of subducted basaltic crust is unlikely to occur beneath
southern Kamchatka, there are evidences of sediment-melt chemical signature in
Kamchatka, although this contribution is small (< 1%) compared with other volcanic
arcs (Park et al., 2002).
- 115 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING PROCEDURE
A system of 2D Navier-Stokes equations and the 2D steady state heat
transfer equation are solved for the south Kamchatka cross section (Fig. 1) using
the numerical scheme proposed by Manea et al. (2004b). The strong temperaturedependence of viscosity imposed in the present modeling, corresponding to
diffusion creep of olivine, has the following form:
η = η0 ⋅ e
 Ea  T0  
⋅  −1 

 R ⋅T0  T  
(1)
where:
η
-mantle wedge viscosity (Pa s),
20
η0 -mantle wedge viscosity at the potential temperature T0 (10 Pa s),
T0 -mantle wedge potential temperature (1,450ºC),
Ea -activation energy for olivine (300 kJ/mol) (Karato and Wu, 1993),
R
-universal gas constant (8.31451 J/mol ºK),
T
-temperature (ºC).
A finite-element grid extends from 25 km seaward of the trench up to 375
km landward of it, and consists of 12,000 triangular elements with higher resolution
in the tip of the wedge (Fig. 2). A benchmark with various grid resolution of the
present numerical scheme (Manea et al., 2004b) confirms that a numerical error of
less than 5ºC is introduced in the thermal models when 12,000 triangles are used.
The lower edge of the grid follows the shape of the subducting plate upper surface
at a constant distance of 100 km (Fig. 3). The model consists of five thermostratigraphic units as follows: upper continental crust, lower continental crust,
oceanic lithosphere, oceanic sediments, and mantle wedge.
- 116 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
A summary of the thermal parameters used is presented in Table 1
(compilation from: Peacock and Wang, 1999; Smith et al., 1979; Vacquier et al.,
1967). The continental crust in Kamchatka is divided into two layers: the upper
crust (0 - 15 km depth) and lower crust (15 - 35 km depth). These depths are
consistent with values inferred from 1D tomographic inversion by Gorbatov et al.
(2000). A recent paper of Levin et al. (2002) shows a Moho depth range of 30 - 40
km across the entire Kamchatka peninsula. In this chapter we used a constant
Moho depth of 35 km. The shape and dip of the subducting plate beneath the
volcanic arc are constrained by earthquake hypocenter distribution. A 1.5 km-thick
sediment layer is included in the model (Dickinson, 1978; Selivestrov, 1983).
The upper and lower boundaries are maintained at constant temperatures of
0ºC at surface and of 1,450ºC in the asthenosphere, respectively (Fig. 3). The left,
landward vertical boundary condition is defined by an 18.5ºC/km thermal gradient
for the continental crust. Below the 35 km depth, the left boundary condition is
represented by a low thermal gradient of 5.5ºC/km down to the depth of 180 km.
Beneath 180 km depth no horizontal conductive heat flow is specified. Underneath
the Moho (35 km), for the left boundary, corresponding to the mantle wedge, zero
traction is assumed. At the intersection between the subducted slab and the left
boundary, the velocity of the subducting slab is assumed.
The right, seaward boundary condition is a one-dimensional geotherm
calculated for the oceanic plate by allowing a half-space to cool from zero age to
the oceanic plate age at the trench. This geotherm is obtained using a timedependent sedimentation history (Wang and Davis, 1992) and assuming a
constant porosity-depth profile of the sediment column with a uniform sediment
thickness of 1.5 km at the trench.
Since the Stokes equations are applied only in the mantle wedge, the region
between Moho and the slab surface is entirely involved in the flow induced by the
subducting slab. This is consistent with the conclusion of Levin et al. (2002),
suggesting that the mantle lithosphere beneath Kamchatka is actively deforming.
- 117 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
In terms of displacement, the velocity of the oceanic plate is taken with
respect to the continental plate. Thus the convergence rate of 7.8 cm/year between
the PAC and North American plates is used for the KSZ (Renkin and Sclater,
1988). Although we assume the motion of the subducting slab as parallel to the
dip, the slab can have an additional downward directed velocity component due to
slab bending or slab rollback. On the other hand, slab can also have upward
velocity component in case of swallowing of the subduction angle with time due to
the slab unbending. In our models we are dealing with a steady state model and
therefore, do not consider additional velocity component due to the slab
bending/unbending. The PAC age at the trench is 72 Myr according the estimation
of Gorbatov and Kostoglodov (1997).
We consider two different models: the first has constant viscosity
(isoviscosity) in the mantle wedge and the second one has strong temperaturedependent viscosity (diffusion creep of olivine). In the second model, the system of
equations becomes strongly nonlinear. To deal with this difficulty Picard iterations
are applied. In order to achieve a convergent solution a cut-off viscosity of 1024 Pa
s fortemperature less than 1,000 ºC is used.
A long-term continuous sliding between the subducting and continental
plates along the thrust fault should produce frictional heating. We introduced into
the models a small degree of frictional heating using Byerlee’s friction law (Byerlee,
1978). Frictional heating cease at a depth of 35 km, where the oceanic plate and
the mantle wedge came into first contact (Fig. 3). We impose this depth limitation
of frictional heating because the contact between Moho (35 km) and the slab
surface represents a maximum extent where interplate earthquakes might occur.
The tip of the mantle wedge is subject of mantle serpentinization which has a
completely ductile behavior and therefore decoupling the subducting and overriding
plates. The location of serpentinized mantle wedge tip is critical because it controls
the down-dip extension of the interplate earthquakes (Manea et al., 2004a).
Volumetric shear heating is calculated as follows:
- 118 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Qsh =
τ ⋅v
w
(1)
,
where:
Qsh
-volumetric shear heating (mW/m3),
τ
-shear stress:
σn
-lithostatic pressure (MPa),
λ
-the pore pressure ratio, (PPR-the ratio between the hydrostatic and
τ = 0.85 ⋅ σ n ⋅ (1 − λ ) for σ n ⋅ (1 − λ ) ≤ 200MPa
,
τ = 50 + 0.6 ⋅ σ n ⋅ (1 − λ ) for σ n ⋅ (1 − λ ) > 200MPa
lithostatic pressures. λ <= 1. The maximum value, λ = 1, means no frictional
heating),
v
-convergence velocity (7.8 cm/year),
w
-the thickness of a thin element layer (500 m) along the plate-interface,
where frictional heating is formulated as body-heat source.
Based on the velocity field obtained in the case of temperature dependent
viscosity, a dynamic model for the blob tracers is applied. A blob moves under the
action of drag, mass, and buoyancy forces in the mantle wedge stationary velocity
field generated in the previous model. The description of the modeling approach is
given in (Manea et al., 2004b). The trajectories of positively buoyant blobs (∆ρ =
200 kg/m3) with diameters between 0.4 and 10 km are calculated for different
values of wrapping viscosity, ηw (1014 - 2.1017 Pa s). Very low wrapping viscosity
around the blobs might be a consequence of viscous heating between the
surrounding mantle and the blob (Gerya and Yuen, 2003). The total rise time that
the blobs require to reach the base of the continental crust is also estimated.
During migration through the mantle wedge the blobs are heated first
because of the inverted thermal gradient, and then cooled (normal thermal
gradient) before approaching the base of the continental lithosphere. A blob is
- 119 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
assumed to be heated/cooled by conduction only. The following conduction
equation is used to model blob thermal history:
Cp ⋅
∂T
+ ∇ ⋅ (−k ⋅∇T ) = 0
∂t
(2)
where:
Cp
-thermal capacity 3.3 (MJ/m3 °K),
T
-temperature (°C),
t
-time (Myr),
k
-thermal conductivity 3.1 (W/m °K),
The equation is solved numerically inside the blob, with boundary conditions
of mantle wedge temperature taken from the thermal modeling. We used 3,000
triangle elements to solve equation (2) inside the spherical blobs. The trajectory of
the blobs and the time steps are taken from the dynamic model for the blob tracers
described above.
Finally, a tomographic image is obtained using the thermal models and
temperature dependence of seismic-wave velocities from (Karato, 1993).
The
seismic-velocity perturbations are calculated relative to the PREM model of
Dziewonski and Anderson (1981). The following equation (Karato, 1993) is used:
−1
H
∂ ln V ∂ ln V0 Q p
=
−(
⋅
)
π R ⋅T 2
∂T
∂T
(3)
where:
V
-velocity (km/s),
V0
-the reference velocity (Dziewonski and Anderson, 1981),
T
-temperature (°C),
- 120 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
∂ ln V0
∂ ln V0
= −α ⋅
∂T
∂ ln ρ
(3.1)
with:
α
pressure dependence of thermal expansion (Stacey, 1977),
∂ ln V0
∂ ln ρ
from (Chopelas, 1992).
Qp-1 - seismic anelasticity for compressional waves in peridotite (spinel lhezoite)
determined by Sato et al., 1988, 1989 as follows:
[ g '⋅(
Q p = Q pm ⋅ e
Tm
− a )]
T
(3.2)
where:
Qpm - is the Qp at solidus temperature and it has the following form:
(3.3)
Qpm=Q1+P/P0,
with:
Q1 =3.5
P0
=73 MPa
P
-lithostatic pressure (MPa),
g’ = 6.75, a = 1.00 for Tm/T<1,
g’ = 8.47, a = 1.00 for 1<Tm/T<1.08,
- 121 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
g’ = 13.3, a = 1.03 for Tm/T>1.08,
Tm -dry peridotite solidus (Wyllie, 1979),
H
-activation enthalpy for olivine: 500 kJ/mol (Karato and Spetzler, 1990),
R
-universal gas constant (8.31451 J/mol.K),
T
-temperature (ºC).
- 122 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING RESULTS
Thermal models
The thermal models that correspond to isoviscosity and temperaturedependent viscosity are presented in Fig. 4 and Fig. 5. The isoviscous mantle
wedge model predicts temperatures of ~ 950 ºC in the asthenosphere, beneath the
volcanic chain indicating that, at least for dry olivine, melting should not occur. The
geotherm of the oceanic plate surface does not intersect the solidus (Fig. 6-A)
neither for basalt nor for wet sediments (Fig. 6-B) indicating that the oceanic plate
and the subducted sediments do not undergo melting. The mantle-wedge velocity
field has a rather small back flow of ~ 3 cm/yr which is responsible for the low
temperature in the wedge (Fig. 7-A). A small amount of frictional heating (a pore
pressure ratio λ = 0.96 or an effective friction coefficient µ = 0.034 and an average
shear stress along the thrust fault τ ~ 14 MPa) is added at the contact between the
subducted slab and the overriding plate. Although the surface of the slab beneath
the mantle wedge is heated by more than 150ºC, this is not sufficient to melt the
slab surface (Fig. 6-A). It is evident that this simple model with the isoviscous
mantle wedge cannot create any source of the volcanic material.
Applying temperature-dependent viscosity in the mantle wedge produces an
important increase of the temperature beneath the volcanic front. The maximum
temperature then rises to more than 1,300ºC (Fig. 5). The mantle wedge viscosity,
η0, at a potential temperature, T0 (1,450ºC), might be between 1017 Pa s and 1020
Pa s. A benchmark for the numerical scheme applied in this study (Manea et al.,
2004b) shows a very small variation in temperature in the wedge of ∆T < 15ºC for
∆η0 =1,000 Pa s (from 1017 Pa s up to 1020 Pa s). Therefore, a unique value for the
mantle wedge viscosity at a potential temperature T0 of η0 = 1020 Pa s is used,
since its effect is negligible on the overall wedge thermal structure.
- 123 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The velocity field (Fig. 7-B) has back-flow velocities of ~ 7.5 cm/yr, which
are responsible for the higher temperature in the tip of the wedge. Without frictional
heating, the geotherm of the subducting plate surface does not intersect the
solidus neither for basalt nor for sediments (Fig. 6-A) therefore melting of the
subducted sediments and basaltic oceanic crust does not occur for this model. The
frictional heating fraction corresponding to λ = 0.96 might cause fluid-saturated
melting of subducted sediments at a relatively shallow depth of ~ 35 km (Fig. 6-B).
The same amount of frictional heating (λ = 0.96) is not sufficient to melt the basaltic
subducted crust. For greater amounts of frictional heating (i.e. λ = 0.95) melting of
basalt oceanic crust might be possible, but this is inconsistent with the lack
adakites volcanism in southern Kamchatka.
Metamorphic sequences and dehydration within the descending oceanic
crust
The estimated variation of wt% H2O content with depth along the subducting
plate for both models, with isoviscosity and temperature-dependent viscosity, with
and without frictional heating, are presented in Fig. 8 and Fig. 9. The isoviscous
thermal model without frictional heating (Fig. 8-A) shows a fairly simple
metamorphic structure: from Lawsonite - Blueschist facies, the oceanic crust enters
at a depth of ~ 25 km into the stability field of Jaedite – Lawsonite - Blueschist;
from ~ 55 km depth to ~ 85 km depth, the metamorphic facies is represented by
Lawsonite - Amphibole - Eclogite. With a small amount of Zoisite - Eclogite from ~
85 to ~ 100 km depth, the oceanic crust loses completely its hydrous phase
entering finally into the Diamond - Eclogite stability field. Intensive dehydration
occurs during these phase changes, more than 5 wt% H2O being released into the
overlying mantle up to a depth of ~ 100 km (see Fig. 8-A – inset). The same
isoviscous thermal model, but with frictional heating included, predicts a more
complicated metamorphic structure (Fig. 8-B). Up to the contact between the Moho
and the subducted slab, the oceanic crust is represented by Lawsonite - Blueschist
- 124 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
(up to ~ 25 km depth), Epidote - Blueschist (25 - 28 km depth) and Epidote Amphibolite (28 - 35 km depth) facies. Deeper, the metamorphosed structure is
characterized by Eclogite - Amphibole (35 - 40 km depth), Zoisite - Amphibole Eclogite (40 - 80 km depth), Zoisite - Eclogite (80 - 100 km depth) and from ~ 100
km by Diamond - Eclogite facies. Again, rigorous dehydration occur all the way
through these phase changes, more than 5 wt% H2O being released into the
overlying mantle up to a depth of ~ 100 km (see Fig. 8-B – inset).
The thermal model with temperature-dependent viscosity and without
frictional heating reveals a similar metamorphic pattern as the isoviscous model
(without frictional heating): Lawsonite - Blueschist, Jaedite – Lawsonite Blueschist, Jadeite - Epidote - Blueschist, Amphibole - Eclogite, Zoisite - Eclogite
and Diamond - Eclogite (see Fig. 9-A – inset). Strong dehydration (more than 5%wt
H2O released) of wedge peridotite up to ~ 100 km depth is suggested by this
model. The last thermal model proposed in this study, with temperaturedependence of viscosity and frictional heating (λ = 0.96 or µ = 0.034) exposes a
slightly different and more complicated metamorphic arrangement along the
oceanic crust: Lawsonite - Blueschist, Greenschist, Epidote - Amphibolite, Eclogite
- Amphibole, Garnet - Amphibolite, Zoisite - Amphibole - Eclogite, Zoisite Eclogite, Coesite - Eclogite and Diamond - Eclogite facies (see Fig. 9-B - inset).
Hydration of mantle wedge peridotite is likely to occur up to ~ 90 km depth and
melting of fluid-saturated oceanic sediments at shallower depth of ~ 40 km is
suggested by this last thermal model (Fig. 5-B and Fig. 6-B).
Vertical thermal profiles (A-A’) beneath the volcanic front (190 km from the
trench) through the mantle wedge (Fig. 4 and Fig. 5) illustrate that for the models
with isoviscosity, mantle peridotite melting is not possible, while for both strong
temperature-dependent viscosity models melting of wet peridotite is likely to occur
(Fig. 6-B).
- 125 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Diapiric ascent of melted material constrained by geochemical tracers
The thermal model with temperature dependent viscosity and frictional
heating confirms the possibility of melting of fluid-saturated sediments and
hydrated mantle wedge peridotite (Fig. 6-B and Fig. 6-A). A strong dehydration flux
from the slab surface might lower the solidus of mantle wedge allowing the
occurrence of partial melt of hydrated peridotite in the vicinity of the slab surface (<
10 km from the slab surface - see Fig. 9). Using the wet peridotite solidus from
Mysen & Bottcher (1975) (i.e. 800°C at ~ 80 km) the wet peridotite solidus is even
closer to the slab surface (~ 5 km).
Hydrated peridotite with a density lower than the asthenospheric density can
develop buoyant plumes (e.g., Gerya and Yuen, 2003). Gerya and Yuen (2003)
and Manea et al. (2004b) suggested that these blobs might be lubricated by a very
low wrapping viscosity due to viscous heating against the surrounding mantle
wedge. The source of the wrap is melted material coming from the subducting slab,
including melted subducted sediments. Wrapping viscosities down to 1014 Pa s
control diapiric ascent in the wedge. The surface geotherm of the subducted slab
intersects the dehydration melting solidus at ~ 90 km (Fig. 6-A).
The blobs are not necessarily melts; they might be compositionally
instabilities at the slab-wedge interface (Gerya and Yuen, 2003). Actually the blobs
are not restricted to be melted, they need only positive buoyancy in order to detach
from the slab surface. The origin of this positive buoyancy (low density) might be
compositional and/or thermal. Consequently we selected an initial point to calculate
blob trajectories at a depth of ~ 110 km on the slab surface, just below the EVF.
The blob trajectories are shown in Fig. 10 for different wrapping viscosities
(1014 - >1017 Pa s) and blob diameters (0.4 - 10.0 km). Very low viscosity is
essential for the blob to rise to the base of continental crust. Extremely low
viscosity (1014 - 1017 Pa s) is also essential to explain the high pressure and ultrahigh pressure metamorphic rocks exhumation from great depths (Burov et al.,
2000).
- 126 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
For blobs of 4 km diameter (Fig. 10-A) and wrapping viscosity η > 3·1016 Pa
s, the drag force is predominant and the blob cannot rise. For lowered viscosity the
drag force is less significant and at the depth of ~ 175 km the blob intercepts the
mantle wedge back flow, which returns it toward the tip of the wedge. Finally the
blob rises and impinges on the continental crust after ~ 8 Myr. For lower viscosity
(< 1016 Pa s) blobs rise faster (Fig. 11) and impact at approximately the same point
below the volcanic chain.
The larger the blob size the less time is required to reach the continental
crust (Fig. 11). The buoyancy force of large blobs becomes more dominant than
the drag force and this yields a substantially upright trajectory (Fig. 10-B). At a
wrapping viscosity of 1015 Pa s blobs with a diameter of less than ~ 0.8 never rise
up to the continental crust. Blobs smaller than ~ 1 km reach the crust at some
distance (up to 10 km), depending on the blob size. For a viscosity of > 10 17 Pa s,
blobs with diameters ≥ 8 km are able to escape from the downward flow in the
mantle and accumulate at the base of the continental crust.
An important constraint on the magma transport time through mantle wedge
comes from U-series isotope disequilibria,
10
Be and fluid soluble trace elements as
Th, Sr and Pb isotope studies.
U-series isotope disequilibria come from the mobility of U in aqueous fluids
under oxidizing conditions. On the other hand Th in not mobile, therefore the
timescale of such U-Th disequilibria might be used to infer the fluid-transfer and
melt generation in the mantle wedge before surface eruptions.
Turner and Hawksworth (1997) proposed a very rapid ascent of magma (~
1000 years) from the place where might be formed (near the slab surface) toward
the earth surface through a channel network. Regelous et al. (1997) infer the
magma transport rate from Th, Sr and Pb isotope data. They analyzed data from
Tonga-Kermadec arc lavas, the same as Turner and Hawksworth (1997). The final
conclusion of Regelous et al. (1997) is that magmas are erupted at the surface <
350 kyr after the melts are generated in the mantle wedge. These results are quite
different, clearly showing that this issue with U-Th isotope disequilibria and Th, Sr
- 127 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
and Pb isotope data used to infer the magma transport rates across the mantle
wedge is an on going debate.
The cosmic ray produces
sediments. Studies of
10
10
Be, which is subducted together with the ocean
Be show the Be can be transported from the trench
through the mantle wedge and finally erupted in surface lavas over a period of ~ 7
Myr (Brown et al. (1982); Morris et al. (1990)). With a subduction velocity rate of
7.8 cm/yr it took ~ 1 Myr for the sediments to arrive at ~ 100 km depth beneath the
EVF. Then the remaining ~ 6 Myr represents the residence time for Be in the
mantle wedge.
The U series and
10
Be studies show contradictory estimates of the magma
transport times through the mantle wedge, suggesting variable transport rates for
different elements and/or real differences in transport times probably due to a
variable melted volume generated in the wedge.
Whether we consider in our models the results of Regelous et al. (1997),
(350 kyr for magma transport), then from Fig. 11 - inset can be seen that a magma
transport mechanism as buoyant blobs still represents a reliable mechanism. The
blobs with the size of less than 10 km can reach the continental crust in less than
350 kyr for wrapping viscosities less than ~ 1016 Pa s.
Assuming a residence time of magma of ~ 6 Myr (from
10
Be studies), then
buoyant blobs with diameters of 0.4-10 km are allowed to travel through mantle
wedge toward the Moho for a very wide range of wrapping viscosities of 10 14 - >
1017 Pa s.
Thermal history of the blobs
Heat transfer accompanies the journey of cold buoyant blobs through the
mantle wedge. A temperature-dependent viscosity thermal model was used for a
10-km diameter buoyant blob. The thermal history of this blob, which reaches the
Moho in 2.6 Myr, is presented for nine time periods in Fig. 12-A relatively cold blob
(~ 800ºC) initiates its voyage through the inverted thermal gradient. Its top is
- 128 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
heated up to ~ 950ºC by the surrounding mantle (Fig. 12-A). After ~ 0.33 Myr, the
blob has moved down few km from its initial position, being dragged by the
vigorous mantle wedge flow. The top hot region grows in size and temperature (~
1,170ºC) while the cold core shrinks and becomes warmer (~ 810ºC) (Fig. 12-B).
The downward trajectory of the blob continues, and after ~ 0.66 Myr inner
temperatures between 1,000ºC and 1,290ºC are expected (Fig. 12-C). From this
point the blob starts to rise, being close from the maximum wedge temperature
after ~1 Myr, when the temperatures inside the blob are from 1,180ºC up to
1,330ºC. Beyond this point the blob rises in a normal thermal gradient and after ~
2.6 Myr reaches the base of the continental crust with a > 950ºC hot core while the
surrounding mantle has ~ 800ºC (Fig. 12-A-i). Assuming that the main composition
of the blob is peridotite and using the wet peridotite solidus from Mysen and
Boettcher (1975) (e.q. 850ºC at 35 km depth) the main blob’s volume is not
solidified when touches the Moho due to hot core temperature above 950ºC (Fig.
12 and Fig. 13).
Neither viscous heating around the diapir nor thermal convection inside are
incorporated in this model. For a blob with d = 10 km diameter, with a minimum
blob viscosity µ = 1017 Pas and a maximum thermal contrast of ∆T = 350ºC (see
Fig. 12-A-b), the Rayleigh number is Ra
max
ρb ⋅ g ⋅α ⋅ ∆T ⋅ d 3
= 1,050 (other
=
µ ⋅k
parameters for Ramax calculation are: ρb = 3,000 kg/m3, g = 10 m/s2, α = 10-5 °C-1, k
= 10-6 m2/s). Although Ramax exceeds Racr = 660 necessary for convection to begin
in fluid layers heated from below (or above) (Turcotte and Schubert, 2002), in the
present study we do not incorporate the effect of thermal convection on the blob’s
thermal structure. Every parameter in the Rayleigh number is pretty well known
except blob’s viscosity, which can vary by orders of magnitude. Future
investigations will focus on the effect of viscous heating and thermal convection
over the thermal history of buoyant blobs.
- 129 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Velocity anomaly estimation from thermal modeling
The tomographic image computed using the thermal model with
temperature dependent viscosity is presented in Fig. 14-A. The high temperature in
the mantle wedge beneath the volcanic chain (> 1,300ºC) produces a strong
negative velocity anomaly up to –7% (Fig. 13) (relative to PREM). The cold
subducting slab produces a positive velocity anomaly up to +4%.
The tomographic image estimation from isoviscous thermal models revealed
very low amplitude velocity perturbations in the mantle wedge, because of the
thermal structure mainly controlled by the left boundary condition which is
consistent with the PREM model.
The procedure applied to estimate the tomography anomaly from a thermal
model applying Karato (1993) uses the dry solidus for peridotite (Tm). As is shown
in Fig. 6-B, the wedge temperature is well above the wet solidus for peridotite but is
below the dry solidus, therefore no partial melting effect is included in this
tomography estimation. The thermal models with strong temperature dependence
of viscosity show a temperature of ~ 1,300°C at ~ 90 km depth (see Fig. 6-B). This
corresponds to ~ 90% of the dry peridotite solidus. Tomography anomalies are
usually interpreted as indicating a partially molten asthenosphere, but Sato et al.
(1989) shows that this may reflect instead a hot solid asthenosphere where the
temperature approaches 90% of the dry peridotite solidus. Experimentally anelastic
properties of peridotite determined by Sato et al. (1989) illustrate that the
attenuation mechanism of peridotite might be the weakness (or “softening”) of grain
boundaries at high temperature below the solidus. This might be the case in
southern Kamchatka too, since the wedge temperatures come close to ~ 90% of
the peridotite dry solidus.
- 130 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
DISCUSSION AND CONCLUSIONS
Numerical models of steady-state temperature and velocity fields in the
mantle wedge of the Kamchatka subduction zone are developed using the
numerical scheme of Manea et al. (2004b). Based on this, a dynamic model of
buoyant blob migration in the mantle wedge velocity field is developed. The
thermal history of a 10 km blob, which moves up through the mantle wedge
thermal field is predicted. Finally, the thermal structure of the mantle wedge is used
to estimate the seismic P-wave velocity anomalies (referenced to PREM)
associated with subduction of the Pacific plate beneath Kamchatka. The velocity
anomalies estimates are compared with a seismic tomography image inferred from
P-wave arrivals for the same cross-section (Gorbatov et al., 1999).
Four different models are considered, the first two with isoviscosity in the wedge
(with and without frictional heating) and the second two with the temperaturedependent viscosity (with and without frictional heating). Both type of thermal
models (isoviscous and with the temperature dependent viscosity), show a velocity
inflow-outflow in the mantle wedge beneath Moho and the slab surface. This type
of flow might induce elastic anisotropy in the wedge peridotite, which is consistent
with the trench normal strike of the inferred anisotropic fast axes in southern
Kamchatka (Levin et al. 2002).
The isoviscous thermal models do not predict any melting in the
asthenosphere beneath the volcanic chain (Fig. 4). With a temperature > 1,300 ºC
(Fig. 5), the model with temperature-dependent viscosity in the mantle wedge
shows a significant increase in temperature beneath the volcanic arc. Two different
sources of melt are possible: sediments and wedge peridotite beneath the volcanic
front). A small amount of frictional heating (a pore pressure ratio λ = 0.96 or an
effective friction coefficient µ = 0.034 and an average shear stress along the thrust
fault τ ~ 14 MPa) is added at the contact between the subducted slab and the
overriding plate (Fig. 6-A). For larger amounts of frictional heating (i.e. λ = 0.95)
- 131 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
melting of basalt oceanic crust might become possible, but this is inconsistent with
the lack adakites volcanism in southern Kamchatka.
Partial melting of peridotite in subduction zones is initiated by an influx of
fluids derived from the metamorphosed slab and sediments (Tatsumi, 1986; Davies
and Stevenson, 1992). Major dehydration of the basaltic oceanic crust (> 5 wt%
H2O release) occurs just below the volcanic chain up to a depth of ~100 km (Fig. 8
and Fig. 9). Despite the variation in parameters and model input, dehydration of the
top of the slab is complete in all models at basically the same depth. This depth
estimation is rather robust and does not depend on model parameters. The H2O
contents in the phase diagrams for mafics and harzburgite (Hacker et al., 2003)
represent the maxima. Whether such H2O contents are reached depends on presubduction alteration and fluid flow in the subducted slab. Experimental studies of
fluid-saturated peridotite show a solidus as low as ~ 800 ºC at pressures between
2-3 GPa (Mysen and Boettcher, 1975). As a result, just above the subducting slab
(~ 5 km) a layer of melted peridotite might exist, as is the case for the temperaturedependent viscosity thermal models in the present study. The existence of such a
melted layer has been suggested by Okada (1979). He deduced a low-velocity
layer in the vicinity of the mantle wedge-slab interface from the efficient conversion
of ScS and ScSp phases. Gerya and Yuen (2003) demonstrate that RayleighTaylor instabilities can develop and rise up from the top of cold subducting slabs.
They also suggest that plumes detached from the slab might be lubricated by
partially melted, low-viscosity material from the subducted crust and hydrated
mantle.
Modeling of blob motion in the mantle wedge viscous flow field induced by
the subducting slab, shows that this simple approach may shed light on the origin
of the volcanism beneath south Kamchatka. Two parameters control the
trajectories of blobs rising from the slab: the diameter of the blob and the wrapping
viscosity. Very low values of wrapping viscosity (1014 - 1017 Pa s) are necessary to
counter out the drag and buoyancy forces, which is critical for the blob to rise.
- 132 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Blob rise time decreases nonlinearly as its diameter increases and the
wrapping viscosity diminishes (Fig. 10). The time required for a blob of 10 km
diameter to rise from the slab surface up to the continental crust varies from <
2,000 yr up to 10 Myr for viscosities between 1014 Pa s and > 1017 Pa s
respectively. For a diameter of ≥ 8 km and a viscosity of ≥ 1017 Pa s blobs rise
upward until they reach the base of the continental lithosphere. This low value of
viscosity is in the lowermost range experimentally determined for olivine at the
upper mantle pressures and temperatures (Hirth and Kohlstedt, 2003).
Dynamic models of blob trajectories in the mantle wedge velocity field
shows that “fast” trajectories terminate at the same location on the base of the
continental lithosphere (Fig. 10), while the final points of “slow” trajectories, which
are more common for the blobs of smaller size (< 1 km), are dispersed.
An important constraint on the magma transport time through mantle wedge
comes from U-series isotope disequilibria,
10
Be and fluid soluble trace elements as
Th, Sr and Pb isotope studies. Turner and Hawksworth (1997) proposed a very
rapid ascent of magma (~ 1,000 years or ~ 60 m/yr) from the slab surface toward
the earth surface through a channel network. On the other hand, Regelous et al.
(1997) infer a magma transport rate of < 350 kyr (or ~ 17 cm/yr) from Th, Sr and
Pb isotope data. These results are quite different but in a recent study for U-Th-PaRa disequilibria for Kamchatka of Dosseto et al. (2003), is discussed the existence
of a dynamic melting model which does not require a high upwelling velocity (i.e. 1
m/yr) within the mantle wedge. A more contradictory conclusion comes from
10
Be
studies, which show that residence time for Be in the mantle wedge is ~ 6 Myr
before the surface eruption.
The U-Th-Pa-Ra, Th, Sr and Pb isotope disequilibria and
10
Be studies show
very contradictory estimates of the magma transport times through the mantle
wedge from < 350 kyr up to 6 Myr.
Such high variability in magma transport
suggests a very variable transport rates for different elements and/or real
differences in transport times probably due to a variable melted volume generated
in the wedge. Apart of its inherent simplicity, the advantage of a magma transport
- 133 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
model using buoyant blobs proposed in this chapter is its ability to cover the wide
range of residence magma time inside the mantle wedge (Fig. 11).
The proposed positively buoyant blobs might have a complex composition of
melted saturated peridotite and melted sediments. An H2O-rich component (likely
with sediments) resulting from dehydration penetrates overlying peridotites.
Ascending into the hotter mantle, this material passes the wet solidus of peridotite
(see Fig. 14). Partial melting starts and the buoyant blobs begin to move through
the strong mantle wedge flow. Finally the blobs reach the base of the lithosphere
and form a magmatic chamber beneath the volcanic chain.
Levin et al. (2002) proposes a scenario where the anisotropy in the top layer
(~ 15 km beneath Moho) of the mantle wedge beneath EVF in southern Kamchatka
is generated by melt lenses and/or sheeted diapirs. In this model, the anisotropy
might be generated without the presence of a mantle wedge inflow beneath Moho.
This model is in good agreement with the flow model presented in Fig. 7-B, where
very small inflow velocities are obtained from the thermal model with strong
temperature dependence of olivine. Moreover, the magma propagation model with
buoyant blobs shows an accumulation area of melted blobs just beneath the
volcanic arc at the base of Moho (Fig. 10).
One important control regarding the reality of buoyant blobs comes from a
study of deformed peridotite xenoliths from Avachinscky volcano (Fig. 1) (Graybill
et al., 1999). In this paper the xenoliths strains do not indicate a shear induced by a
corner flow, rather they seems to belong from individual diapirs traveling through
the mantle wedge.
The thermal evolution of the blobs was investigated by applying the heat
conduction equation. Thermal convection inside a blob of 10 km diameter is not
likely to occur because the Rayleigh number is very small (~ 2.7 for a viscosity of
1017 Pa s and a maximum thermal contrast of 300ºC). Therefore, in the present
study, the blob is heated/cooled only by conduction. After about 1 Myr, the cold
blob (~ 800ºC) moves toward the hotter region of the wedge where a temperature
of more than 1,300ºC is estimated (Fig. 12 and Fig. 14). After being heated the
- 134 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
blob moves upwards toward the base of the continental lithosphere where it arrives
with a hot core (> 900ºC). According to the fluid-saturated peridotite solidus of
Mysen and Boettcher (1975), most of the 10 km blob (T > 900ºC at 1 GPa) is
melted when arrives at the Moho base (Fig. 12-A-i and Fig. 13). The blobs may
carry fluid-saturated melts (i.e. melted peridotite) and sediments from the
subducted slab to the base of the continental lithosphere and therefore trace
elements and the chemical signature finally might reach the earth’s surface through
volcanic eruptions. Although Fedotov and Masurenkov (1991) classified the active
volcanoes in Kamchatka as calc-alkaline, Tatsumi et al. (1994) interpreted as
tholeiitic the southern Kamchatka volcanoes. However, Kelemen (1990) shows that
“slow” ascent of a tholeiitic melt through a hot mantle wedge produces calc-alkaline
magmatism at the surface.
The thermal model with temperature-dependent viscosity and frictional
heating is used to estimate a seismic tomography image below southern
Kamchatka. The high temperatures in the mantle wedge beneath the volcanic
chain (> 1,300ºC) produce a strong negative velocity anomaly of up to –7 % (Fig.
14-A) (relative to PREM). On the other hand, the cold subducting slab produces a
positive velocity anomaly up to +4%. Since the seismic tomography of Gorbatov et
al. (1999) has a small uncertainty of ~ 10%, our estimation is in good agreement
with the velocity anomalies obtained by Gorbatov et al. (1999) for a 2D profile
identically located with our 2D cross sections. The shape of our tomographic image
inferred from thermal modeling differs from the P-wave seismic tomography image
of Gorbatov et al. (1999), especially for the continental lithosphere and the
uppermost mantle beneath the volcanic chain (Fig. 14-B). This is likely due to the
fact that the thermal models in this study do not consider the magma transport
effect toward the surface. However, since the resolution of the seismic tomography
is very low (50 km) the position of the low velocity zone beneath the EVF has very
large uncertainties. Nevertheless, good agreement of the velocity perturbation
beneath the volcanic arc (at least in magnitude) between the tomography image
- 135 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
from P-wave arrivals and our estimation from thermal modeling suggests
satisfactory modeling the mantle wedge beneath southern Kamchatka.
- 136 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
REFERENCES
Billen, M.I., and Gurnis, M., 2001. A low viscosity wedge in subduction zones.
Earth Planetary Science Letters, v. 193, no. 1-2, p. 227-236.
Brown, L., Klein, J., Middleton, R., Sacks, I.S., and Tera, F., 1982. 10Be in island
arc volcanoes and implications for subduction. Nature, v. 299, p. 718-720.
Burov, E., Jolivet, E., Le Pourhiet, L. and Poliakov, A., 2000. A
thermomechanical model of exhumation of high pressure (HP) and ultra-high
pressure (UHP) metamorphic rocks in Alpine-type collision belts. Tectonophysics,
v. 342, p. 113-136.
Byerlee, J.D., 1978. Friction of rocks. Pure Applied Geophysics, v. 116, p. 615626.
Chopelas, A., 1992. Sound velocities of MgO to very high compression. Earth and
Planetary Science Letters, v. 114, p. 185-192.
Davies, J.H., and Stevenson,D.J., 1992. Physical model of source region of
subduction zone volcanism. Journal of Geophysical Research, v. 97, p. 20372070.
Davies.J.H., 1999. The role of hydraulic fractures and intermediate-depth
earthquakes in generating subduction-zone magmatism. Nature, v. 398, p. 142145.
Dickinson, W.R., 1978. Plate tectonic evolution of north Pacific rim, in
Geodynamics of the western Pacific. edited by Ueda, S., Murphy, R.W., and
Kobayashi, K., p. 1-19, Japan Scientific Societies Press, Tokyo, Japan.
Dosseto, A., Bourdon, B., Joron, J.L., and Dupré, B., 2003. U-Th-Pa-Ra study
of the Kamchatka arc: New constraints on the genesis of arc lavas. Geochimica et
Cosmochimica Acta, v. 67, no. 15, p. 2857-2877.
Dziewonski, A.M., and Anderson, D.L., 1981. Preliminary reference Earth model.
Physics of the Earth and Planetary Interiors, v. 25, p. 297-356.
- 137 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Eiler, J.M., Crawford, A., Elliott, T., Farley, K.A., Valley, J.W., and Stolper,
E.M., 2000. Oxygen isotope geochemistry of oceanic-arc lavas. Journal of
Petrology, v. 41, p. 229-256.
Fedotov, S.A., and Masurenkov, Yu.P. (Eds.), 1991. Active Volcanoes of
Kamchatka. Nauka, Moscow, Russia.
Gaetani, G.A., and Grove, T.L., 2003. Experimental Constraints on melt
generation in the mantle wedge: in Inside the Subduction Factory, Geophysical
Monograph 138, edited by J. Eiler, AGU, Washington DC, p. 107-134.
Gerya, T.V. and Yuen, D.A., 2003. Rayleigh-Taylor instabilities from hydration and
melting propel ‘cold plumes’ at subduction zones. Earth and Planetary Science
Letters, v. 212, p. 47-62.
Gorbatov, A., and Kostoglodov, V., 1997. Maximum depth of seismicity and
thermal parameter of the subducting slab: general empirical relation and its
application. Tectonophysics, v. 277, p. 165-187.
Gorbatov, A., Kostoglodov, V., Suárez, G., and Gordeev, E., 1997. Seismicity
and structure of the Kamchatka subduction zone. Journal of Geophysical
Research, v. 102, no. B8, p, 17,883-17,898.
Gorbatov, A., Dominguez, J., Suárez, G., Kostoglodov, V., and Gordeev, E.
1999. Tomographic imaging of the P-wave velocity structure beneath the
Kamchatka peninsula. Journal of Geophysical Research, v. 137, p. 269-279.
Gorbatov, A., Widiyantoro, A., Fukao, Y., and Gordeev, E., 2000. Signature of
remnant slabs in the North Pacific from P-wave tomography. Geophysical Journal
International, v. 142, p. 27-36.
Graybill, J., Brandon, M.T., and Kepezhinskas, P.K., 1999. Olivine lattivepreffered orientation in xenoliths from the mantle wedge beneath the southern
Kamchatka volcanic arc (abstract). Eos Trans. AGU, 80, F926.
Hacker, B.R., Abers, G.A., and Peacock, S.M., 2003. Subduction Factory 1.
Theoretical mineralogy, densities, seismic wave speeds, and H2O contents.
Journal of Geophysical Research, v. 108, 10.1029/2001JB001127.
- 138 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Hart, S.R., 1993. Equilibration during mantle melting: a fractal tree model.
Proc.National Academy of Science, USA, v. 90, p. 11,914-11,918.
Hirth, G., and Kohlstedt, D.L., 2003. Rheology of the mantle wedge: in Inside the
Subduction Factory, Geophysical Monograph 138, edited by J. Eiler, AGU,
Washington DC, p. 83-105.
Johnson, M.C., and Plank, T., 1999. Dehydration and melting experiments
constrain the fate of subducted sediments. G-cubed, 1.
Karato, S., and Spetzler, H.A., 1990. Defect microdynamics and physical
mechanisms of seismic wave attenuation and velocity dispersion in the Earth’s
mantle. Review of Geophysics, v. 28, p. 399-421.
Karato, S.I., 1993. Importance of anelasticity in the interpretation of seismic
tomography. Geophysical Research Letters, v. 20, p. 1623-1626.
Karato, S., and Wu, P., 1993. Rheology of the upper mantle: a synthesis. Science,
v. 260, p. 771-778.
Kelemen, P.B., 1990. Reaction between ultramafic rock and fractionating basaltic
magma. I. Phase relations, the origin of calc-alkaline magma series, and the
formation of discordant dunite. Journal of Petrology, v. 31, p. 51-98.
Kelemen, P.B., Rilling, J.L., Parmentier, E.M., Mehl, L., and Hacker, B.R., 2003.
Thermal Structure due to Solid-State Flow in the Mantle Wedge Beneath Arcs: in
Inside the Subduction Factory, Geophysical Monograph 138, edited by J. Eiler,
AGU. Union, Washington DC, p. 293-311.
Klein, E.M., and Langmuir, C.H., 1987. Global correlations of ocean ridge basalt
chemistry with axial depth and crustal thickness. Journal of Geophysical Research,
v. 92, p. 8089-8115.
Kushiro, I., 1990. Partial melting of mantle wedge and evolution of island arc cruz.
Journal of Geophysical Research, 95, 15,929-12,939.
Langmuir, C.L., Klein, E.M., and Plank, T., 1992. Petrological systematics of midocean ridge basalts: Constraints on melt generation beneath ocean ridges. In
Mantle flow and melt generation at mid-ocean ridges, AGU Monograph 71, edited
- 139 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
by J. Phipps Morgan, D.K. Blackman, and J.M. Sinton, p. 183-280, AGU,
Washington DC.
Levin, V., Park, J., Brandon, M., Lees, J., Peyton, V., Gordeev, E., and Ozerov,
A., 2002. Crust and upper mantle of Kamchatka from teleseismic receiver
functions. Tectonophysics, v. 358, p. 233-265.
Manea, V.C., Manea, M., Kostoglodov, Currie, C.A., and Sewell, G., 2004a.
Thermal Structure, Coupling and Metamorphism in the Mexican Subduction Zone
beneath Guerrero. Geophysical Journal International, 158, 775–784 doi:
10.1111/j.1365 - 246X.2004.02325.x.
Manea, V.C., Manea, M., Kostoglodov, V., Sewell, G., 2004b. (in press)
Thermo-mechanical model of the mantle wedge in Central Mexican subduction
zone and a blob tracing approach for the magma transport. Physics of the Earth
and Planetary Interiors.
Molnar, P., and England, P., 1990. Temperatures, heat flux, and frictional stress
near Major thrust faults. Journal of Geophysical Research, v. 95, p. 4883-4856.
Molnar, P., and England, P., 1995. Temperatures in zones of steady state
underthrusting of young oceanic lithosphere. Earth and Planetary Science Letters,
v. 131, p. 57-70.
Morris, J., Leeman, W.P., and Tera, F., 1990. The subducted component in island
arc magmas.: constraints from Be isotopes and B-Be systematics. Nature, v. 344,
p. 31-36.
Mysen, B.O., and Boettcher, A.L., 1975. Melting of a hydrous mantle: I. Phase
relations of natural peridotite at high pressures and temperatures with controlled
activities of water, carbon dioxide, and hydrogen. Journal of Petrology, v. 16, part
3, p. 520-548.
Nichols, G.T., Wyllie, P.J., and Stern, C.R., 1994. Subduction zone melting of
pelagic sediments constrained by melting experiments. Nature, v. 371, p. 785-788.
Okada, H., 1979. New evidences of the discontinuous structure of the descending
lithosphere revealed by ScSp phase. Journal of Physics of the Earth, v. 27, p. S53S63.
- 140 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Park, J., V. Levin, M. T. Brandon, J. M. Lees, V. Peyton, E. Gordeev, and A.
Ozerov, 2002. A dangling slab, amplified arc volcanism, mantle flow and seismic
anisotropy near the Kamchatka plate corner. in Plate Boundary Zones, Stein, S.,
and Freymueller, J. (eds.), AGU Geodynamics Series No. 30, AGU, Washington
DC, p. 295-324.
Peacock, S.M., 1990a. Fluid processes in subduction zone. Science, v. 248
(4953), p. 329-337.
Peacock,
S.M.,
1990b.
Numerical
simulation
of
metamorphic
pressure-
temperature-time paths and fluid production in subducting slabs. Tectonics, v. 9(5).
Peacock, S.M., 1991. Numerical simulation of subduction zone pressuretemperature-time paths: Constraints on fluid production and arc magmatism: in
Phil. Trans. Roy. Soc. London A 335: The Behavior and Influence of Fluids in
Subduction Zones, edited by Tarney, J., Pickering, K.T., Knipe, R.J., and Dewey,
J.F., p. 341-353, London, United Kingdom.
Peacock, S.M., 1996. Thermal and petrologic structure of subduction zones, in
Subduction Zones, Top to Bottom: Geophysical Monograph 96, edited by Bebout,
G.E., Scholl, D.W., Kirby, S.H., and Platt, J.P., p. 119-133, AGU, Washington DC.
Peacock, S.M., and Hyndman, R.D., 1999. Hydrous minerals in the mantle wedge
and the maximum depth of subduction thrust earthquakes. Geophysical Research
Letters, v. 26, p. 2517-2520.
Peacock, S.M., and Wang, K., 1999. Sesimic consequences of warm versus cool
subduction metamorphism: Examples from southwest and northeast Japan.
Science, v. 286, p. 937-939.
Peacock, S.M., (in press) 2002. Thermal structure and metamorphic evolution of
subducting slabs: in Eiler, J. (ed.) The Subduction Factory, AGU Geophysical
Monograph.
Plank, T., and Langmuir, C.H., 1993. Tracing trace elements from sediment input
to volcanic output at subduction zones. Nature, v. 362, p. 739-743.
Regelous, M., Collerson, K.D., Ewart, A., and Wendt, J.I., 1997. Trace element
transport rates in subduction zones: evidence from Th, Sr and Pb isotope data for
- 141 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Tonga-Kermandec arc lavas. Earth and Planetary Science Letters, v. 150, p. 291302.
Renkin, M.L., and Sclater, J.G., 1988. Depth and age in the North Pacific. Journal
of Geophysical Research, v. 93, B4, p. 2919-2935.
Rubin, A.M., 1993. Dikes vs. diapirs in viscoelastic rock. Earth and Planetary
Science Letters, v. 117, p. 653-670.
Sato, H., Sacks, I.S., Takahashi, E., and Scarfe, C.M., 1988. Geotherms in the
Pacific Ocean from laboratory and seismic attenuation studies. Nature, v. 336, p.
154-156.
Sato, H., Sacks, I.S., Murase, T., Muncill, G., and Fukuyama, H., 1989. QPmelting temperature relation in peridotite at high pressure and temperature:
attenuation mechanism and implications for the mechanical properties of the upper
mantle. Journal of Geophysical Research, v. 94, p. 10,647-10,661.
Selivestrov, N.I., 1983. Structure of the junction zone of the Kurile-Kamchatka and
Aleutian island arcs according to data from continuous seismic profiling (in
Russian). Volcanology and Seismology, v. 2, p. 53-67.
Shaw, B.E., 1980. The fracture mechanisms of magma transport from the mantle
to the surface: In: Hargraves, R.B. (Eds.), Physics of Magmatic Processes.
Princeton Univ. Press, Princeton, p. 201– 264.
Smirnov, Ya. B., and Sugrobov, V.M., 1979. Terrestrial heat flow in the KurileKamchatka and Aleutian provinces. I. Heat flow and tectonics (in Russian).
Volcanology and Seismology, v. 1, p. 59-73.
Smirnov, Ya. B., and Sugrobov, V.M., 1980a. Terrestrial heat flow in the KurileKamchatka and Aleutian provinces. II. The map of measured and background heat
flow (in Russian). Volcanology and Seismology, v. 1, p. 16-31.
Smirnov, Ya. B., and Sugrobov, V.M., 1980b. Terrestrial heat flow in the KurileKamchatka and Aleutian provinces. III. Assessments of temperature at depth and
thickness of the lithosphere (in Russian). Volcanology and Seismology, v. 1, p. 1631.
- 142 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Smith, D.L., Nuckels, C.E., Jones, R.L. and Cook, G.A., 1979. Distribution of
heat flow and radioactive heat generation in northern México. Journal of
Geophysical Research, v. 84, p. 2371-2379.
Spiegelman, M., and Kelemen, P.B., 2003. Extreme chemical variability as a
consequence of channelized melt transport. G-cubed, v. 4, no. 7, 1055,
doi:10.1029/2002GC000336.
Stacey, E.D., 1977. A thermal model of the Earth. Physics of the Earth and
Planetary Interiors, v. 15, p. 341-348.
Stolper, E.M., and Newman, S., 1994. The role of water in the petrogenesis of
Mariana trough magmas. Earth and Planetary Science Letters, v. 121, p. 293-325.
Tamura, Y., Tatsumi, Y., Zhao, D., Kido, Y., and Shukuno, H., 2002. Hot fingers
in the mantle wedge: new insights into magma genesis in subduction zones. Earth
and Planetary Science Letters, v. 197, no. 1, p. 105-116.
Tatsumi, Y., Sakuyama, M., Fukuyama, H., and Kushiro, I., 1983. Generation of
arc basalt magmas and thermal structure of the mantle wedge in subduction
zones. Journal of Geophysical Research, v. 88, p. 5815-5825.
Tatsumi, Y., 1986. Formation of the volcanic front in subduction zones.
Geophysical Research Letters, v. 13, p. 717-720.
Tatsumi, Y., Furukawa, Y., Kogiso, T., Yamanaka, K., Yokoyama, T., and
Fedotov, S.A., 1994. Unusual three volcanic chains in the Kamchatka:
Geophysical Research Letters, v. 21, p. 537-540.
Turcotte, D.L., and Schubert, G., 2002. Geodynamics, 2nd edition, Cambridge
University Press, New York.
Turner, S., and Hawkesworth, C., 1997. Constraints on flux rates and mantle
dynamics beneath island arcs from Tonga-Kermadec lava geochemistry. Nature,
v. 389, p. 568-573.
Vacquier, V., Sclater, J.G., and Corry, C.E., 1967. Studies of the thermal state of
Earth: The 21st paper: Heat-flow, eastern Pacific. Bulletin of Earthquake Research
Institute, v. 45, p. 375-393.
- 143 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
van Keken, P.E., Kiefer, B., and Peacock, S.M., 2002. High resolution models of
subduction zones: Implications for mineral dehydration reactions and the transport
of water into deep mantle. G-cubed, v. 3, no. 10, p. 20.
von Hunen, J., van den Berg, A.P., Vlaar, N.J., 2000. A thermo-mechanical
model of horizontal subduction below an overriding plate. Earth and Planetary
Science Letters, v. 182, p. 157-169.
Wang, K. and Davis, E.E., 1992. Thermal effect of marine sedimentation in
hydrothermally active areas. Geophysical Journal International, v. 110, p. 70-78.
Wyllie, P.J., 1979. Magmas and volatile components. American Mineralogy, v.
654, p. 469-500.
Zhao, D., Hasegawa, A., and Horiuchi, S., 1992. Tomographic imaging of P and
S wave velocity structure beneath northwestern Japan. Journal of Geophysical
Research, v. 97, no. 13, p. 19,909-19,928.
Zhao, D., Yingbiao, X, Weins, D.A., Dorman, L., Hildebrand, J., and Webb, S.,
1997. Depth extent of the Lau back-arc spreading center and its relation to
subduction proceses. Science, p. 254-257.
Zhao, D., 2001. Seismological structure of subduction zones and its implications
for arc magmatism and dynamics. Physics of the Earth and Planetary Interiors, v.
3999, p. 1–18.
- 144 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Table 1. Summary of thermal parameters used in the models. (Compilation from
Peacock and Wang, 1999; Smith et al., 1979)
Geological Unit
Oceanic
sediments
Density
(g/cm3)
Thermal
Conductivity
(W/m°K)
Radiogenic
heat
Thermal Capacity
production
(MJ/m3 °K)
(µW/m3)
2.20
1.00 – 2.00*
1.00
2.50
2.70
2.00
1.3
2.50
2.70
2.00
0.2
2.50
3.10
3.10
0.01
3.30
3.00
2.90
0.02
3.30
Upper
continental
crust (0-15 km)
Lower
continental
crust
(15-35
km)
Mantle wedge
and Blob
Oceanic
lithosphere
•
Increases linearly with distance from the deformation front up to a depth
of 10 km.
- 145 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
FIGURE CAPTIONS
Figure 1.
Tectonic settings and position of the modeled cross-section (purple thick
line) in Kamchatka. Orange triangles show the location of trench-side and back-arc
volcanoes. Open, semi-transparent arrows show convergence velocities between
the Pacific and North American plates.
Figure 2.
A grid with 12,000 triangles with ~ 2 km mesh resolution was used to solve
the numerical thermal models. The green spherical mesh was used to solve the
heat transfer equation (2) (see text) inside the blob and contains 3,000 triangles.
Figure 3.
Boundary condition and parameters used in the modeling. The upper and
lower boundaries have constant temperatures of 0ºC and 1,450ºC, accordingly.
The continetal plate is fixed. The left (landward) vertical boundary: 18.5ºC/km
thermal gradient in the continental crust (down to 35 km); between 35 km and 180
km depth the thermal gradient is of 5.5ºC/km; underneath 180 km no horizontal
conductive heat flow is specified. Zero tractions are considered beneath Moho (35
Km), at the boundary, which belongs to the mantle wedge. The right (seaward)
boundary condition is a one-dimensional geotherm for the 70 Myr old oceanic
plate. The PAC plate referred to the North American plate has the convergence
velocity of 7.8 cm/yr. Volumetric shear heating is imposed along the plate interface
up to a maximum depth of 35 km (red dashed line), using the Byerlee's friction law
(Byerlee, 1978).
- 146 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4.
(A). Calculated steady-state thermal field for the isoviscous mantle wedge.
Horizontal dashed line shows the Moho (35 km depth). Thick solid magenta line
denotes the top of the subducting slab. No frictional heating along the thrust zone
is included in this model. A-A’ is the vertical temperature profile in Fig. 6-B.
 = 0.034) along the
(B). The same as A., but frictional heating (λ = 0.96 or µ
thrust zone is included in this model.
Figure 5.
(A). Steady-state thermal field for strong temperature-dependent viscosity in
the mantle wedge. Horizontal dashed line shows the Moho (35 km depth). Thick
solid magenta line denotes the top of the subducting slab. No frictional heating
along the thrust zone is included in this model. B-B’ is the vertical temperature
profile in Fig. 6-B.
(B). The same as A., but frictional heating (λ = 0.96 or µ = 0.034) along the
thrust zone is included in this model.
Figure 6.
(A). Phase diagrams for MORB and maximum H2O contents (Hacker et al.,
2003). Z - Zeolite (4.6 wt% H2O), PP - Prehnite - Pumpellyite (4.5 wt% H2O), PA Pumpellyite - Actinolite (4.4 wt% H2O), G - Greenschist (3.3 wt% H2O), LB Lawsonite - Blueschist (5.4 wt% H2O), EpB - Epidote - Blueschist (3.1 wt% H2O),
EpA - Epidote - Amphibolite (2.1 wt% H2O), JEpB - Jadeite - Epidote - Blueschist
(3.1 wt% H2O), EcA - Eclogite - Amphibole (2.4 wt.% H2O), A - Amphibolite (1.3
wt% H2O), GA - Garnet - Amphibolite (1.2 wt% H2O), Gr - Granulite (0.5 wt% H2O),
GGr - Garnet - Granulite (0.0 wt% H2O), JLB - Jaedite - Lawsonite - Blueschist (5.4
wt% H2O), LAEc - Lawsonite - Amphibole - Eclogite (3.0 wt% H2O), JLTS - Jaedite
- Lawsonite - Talc - Schist, ZAEc - Zoisite - Amphibole - Eclogite (0.7 wt% H2O),
AEc - Amphibole - Eclogite (0.6 wt% H2O), ZEc - Zoisite - Eclogite (0.3 wt% H2O),
Ec - Eclogite (0.1 wt% H2O), CEc - Coesite - Eclogite (0.1 wt% H2O), DEc -
- 147 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Diamond - Eclogite (0.1 wt% H2O). Calculated geotherms: dashed blue line - top of
subducting oceanic crust for isoviscous mantle wedge and no frictional heating;
solid blue line – top of subducting oceanic crust for strong temperature-dependent
viscosity and no frictional heating; dashed pink/red line - top of subducting oceanic
crust isoviscosity in the mantle wedge and frictional heating (λ = 0.96 or µ= 0.034);
solid pink/red line - top of subducting oceanic crust for strong temperaturedependent viscosity and frictional heating (λ = 0.96 or µ = 0.034). The maximum
depth of the stable hydrous phases in the oceanic slab is ~ 90 km in case of
variable rheology in the wedge (and with frictional heating (λ = 0.96 or µ= 0.034)).
(B). Phase diagram for harzburgite, and maximum H2O contents (Hacker et
al., 2003). A - Serpentine - Chlorite - Brucite (14.6 wt% H2O), B - Serpentine Chlorite - Phase A (12 wt% H2O), C - Serpentine - Chlorite - Dunite (6.2 wt% H2O),
D - Chlorite - Harzburgite (1.4 wt% H2O), E - Talc - Chlorite - Dunite (1.7 wt% H2O),
F - Anthigorite - Chlorite - Dunite (1.7 wt% H2O), G - Spinel - Harzburgite (0.0 wt%
H2O), H - Garnet - Harzburgite (0.0 wt% H2O).
Continuous thick yellow line indicates the wet solidus for sediments from
Nichols et al. (1994). Calculated geotherms are the same as in A. The temperature
profiles A-A’ and B-B’ (Fig. 4 and Fig. 5) show that for an isoviscous mantle wedge
thermal structure, melting of wet peridotite is not possible, while for temperaturedependent viscosity melting of wet peridotite beneath the volcanic chain is likely to
occur. The wet and dry peridotite are taken from Wyllie (1979).
Figure 7.
(A). The velocity field in the isoviscous mantle wedge. The maximum inflow
velocity is about 3 cm/yr. The return flow (backflow) is horizontal.
(B). The velocity field with the strong temperature-dependent mantle wedge
viscosity. Note that the maximum velocity of the inflow region is about 7.5 cm/yr
that is comparable with the subducting slab velocity. The velocity field presents a
diagonally upward pattern. Note the very low velocity (< 1 cm/yr) just beneath the
Moho.
- 148 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 8.
(A). Metamorphic facies along the subducting oceanic crust corresponding
to the isoviscous model without frictional heating. The inset represents the variation
of wt% H2O along the subducting crust as function of metamorphic sequence.
More than 5 wt% H2O may be released from hydrous phases in the subducting
slab through continuous dehydration.
(B). The same as for Fig. 8-A but with frictional heating (λ = 0.96 or µ =
0.034). More than 3 wt% H2O may be released from hydrous phases in the
subducting slab through continuous dehydration.
Figure 9.
(A). The same as Fig. 8-A but for the temperature-dependent viscosity
model without frictional heating. More than 5 wt% H2O may be released from
hydrous phases in the subducting slab through continuous dehydration.
(B). The same as Fig. 8-A but for the temperature-dependent viscosity
model with frictional heating (λ = 0.96 or µ= 0.034). More than 3 wt% H2O may be
released from hydrous phases in the subducting slab through continuous
dehydration.
Figure 10.
Blob trajectories in steady mantle wedge flow (Fig. 7-B). The initial point for
all trajectories is at a depth of 110 km on the surface of the subducting slab below
the trench-side volcanic belt.
(A). Blob diameter is fixed at 4 km. The trajectories, corresponding to
different wrapping viscosities of less than 1016 Pa s, have the same final point
below the volcanic chain.
(B). The wrapping viscosity is fixed at 1015 Pa s. The trajectories correspond
to different blob size (0.8 - 10.0 km). Blobs with a diameter less than ~ 0.8 never
rise to the continental crust.
- 149 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 11.
Blob rise time as a function of wrapping viscosity. The curves annotated with
blob diameter show that the rise time decreases for bigger blobs and the lower
viscosity. Blobs with a size of less than 10 km can reach the continental crust in
less than 350 kyr for wrapping viscosities less than 1016 Pa s (see inset).
Figure 12.
(A). The thermal history of a 10-km blob which reaches the Moho in 2.6 Myr
at nine time periods. The wrapping viscosity is 1017 Pa s. The circles represent the
blob cross-section.
(B). The trajectory followed by the blob in a thermal structure for
temperature-dependent viscosity (Fig. 5-B).
(C). The same as B. but zoomed.
Figure 13.
P-T Trajectory of the 10-km diameter blob (see Fig. 12) through the mantle
wedge. Colour disks represent the blob at nine time periods from Fig. 12. The
number inside the disks shows the rising time. Dark blue dashed line represents
wet peridotite solidus from Wyllie (1979). Green dashed line represents wet
peridotite solidus from Mysen and Boettcher (1975) (limited to ~ 3 GPa).
Figure 14.
(A). Velocity anomaly estimation below south Kamchatka inferred from
thermal models with temperature-dependent viscosity (Fig. 5-B). Red and blue
colors reveal the slow and fast velocities according to the vertical scale.
(B). Seismic tomography of south Kamchatka inferred from P-wave arrivals
from Gorbatov et al. (1997).
- 150 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 1
- 151 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 2
- 152 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 3
- 153 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4
- 154 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 5
- 155 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 6
- 156 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 7
- 157 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 8
- 158 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 9
- 159 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 10
- 160 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 11
- 161 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 12
- 162 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 13
- 163 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 14
- 164 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
V. LA SÍSMICIDAD INTRAPLACA Y LOS ESFUERZOS TÉRMICOS EN LA
PLACA DE COCOS DEBAJO DE LA PARTE CENTRAL DE MÉXICO
In review at Tectonophysics
INTRASLAB SEISMICITY AND THERMAL STRESS IN THE SUBDUCTED
COCOS PLATE BENEATH CENTRAL MÉXICO
V.C. Manea1, M. Manea1,V. Kostoglodov1, S.K. Singh1, G. Sewell2
1
Instituto de Geofisica, Universidad Nacional Autónoma de México (UNAM),
México
2
University of Texas, El Paso
ABSTRACT
With a maximum depth extent of ~ 80 km, an important particularity of the
intraslab earthquakes beneath Guerrero, is their exclusively normal fault
mechanism. Based on the recently developed thermal models for Guerrero
subduction zone, the thermal stress due to non-uniform temperature distribution in
the subducting slab is calculated using a finite element method. The calculation
results revealed that the first shallow part of the subducting slab is characterized by
low deviatoric compressional thermal stresses in its central part (~ 0.25 kbars).
Following this part is a section where the stress field changes to an extensional
behavior for the core of the slab with maximum values of ~ 0.4 kbars for the flat
region and ~ 0.75 kbars deeper, where the slab bends into the asthenosphere. An
important characteristic for the central Mexican subduction zone is its shallow
subhorizontal plate contact; significant thermal compressional stresses (~ 0.65
kbars) arising in the upper and lower part of the slab are not consistent with the
normal fault intraplate earthquake focal mechanism. Since the Cocos plate dips
- 165 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
into the asthenosphere at an angle of ~ 20º, pressure forces due to the induced
flow in the mantle wedge are partially balanced by the gravitational body forces. A
net clockwise torque might exist at the hinge point and its value is set in such a
way that the compressional stresses for the flat part of the plate vanish. For the
lower part of the slab, ductile behavior is assumed, the compressional stresses
decreasing exponentially with depth. According to the thermal models for Guerrero,
the temperature range of 700ºC - 800ºC is in good agreement with the maximum
depth extent of the intraplate earthquakes and with the maximum extent of the
tensional stresses in the subhorisontal part of the slab.
Keywords: Thermal stress, Bending stress, Mexican subduction zone, flat
subduction.
- 166 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
INTRODUCTION
It has been accepted for some time that the cold subducted oceanic
lithosphere, entering into the hotter mantle, produces significant thermal stresses
that might generate intraslab earthquakes. Thermal stress distribution has been
proposed as possible explanations for a double fault plane in various subduction
zones as Japan and Kamchatka (Hamaguchi et al., 1983, Gorbatov et al., 1997).
The thermal stress distribution of Hamaguchi et al. (1983) reveals very high
compressional stresses up to 10 kbars for the top and bottom of the slab, and
tensional stresses up to 7.5 kbars for the core of the slab. These high values are
with ~ 2 order of magnitude greater than the stress drop during intraslab
earthquakes (~ 0.1 kbars). These excessive values of the thermal stresses are
related with temperature contrast as high as 1,000°C in the subducting slab. The
bottom of the slab is likely to exhibit a ductile behavior due to high temperature,
and therefore the thermal stresses induced in this region should vanish. The
oversimplified mantle wedge thermal structure of Hamaguchi et al. (1983) is
represented by a constant thermal gradient of 10°C/km up to 100 km depth and
5°C/km up to 150 km depth.
The thermo-mechanical models of the mantle wedge with temperature
and/or stress dependent rheology (Furukawa, 1993; Conder et al., 2002; Van
Keken et al., 2002; van Hunen et al., 2002; Kelemen et al., 2003, Manea et al.,
2004b), show a more complicated temperature distribution, where the temperature
contrast inside the descending slab being strongly controlled by the thermal
structure of the mantle wedge. Using the recently developed 2D thermal models,
with strong temperature dependence viscosity in the mantle wedge, for the
Mexican subduction zone (Manea et al., 2004b), thermal stresses induced in the
subducted slab due to non-uniform temperature distribution, are inferred using a
finite element technique. Bending stresses due to concave/convex shape of the
subducting slab are very difficult to determine (Goto et al., 1985) because they are
very sensitive to the curvature of the plate that have large uncertainties in Mexico.
- 167 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Slab pull of the denser subducting slab and the associated down flexure of the slab
produce important deviatoric tension of up to several 100 MPa. Also, important
stress concentrations and pressure variations (up to some 100 MPa), due to the
shape of the subducting slab along the thrust fault, are likely to appear in this
region (Manea et al., 2004a). The state of stress in the slab depends also on
whether the thrust fault is locked or unlocked, producing variations of the slab
interior stress. Although the ridge push might be an order of magnitude smaller
than the trench pull (Turcotte and Schubert, 2002), the resistive forces
encountered by the descending oceanic lithosphere into the mantle may
compensate this difference.
The transmitted ridge push, slab pull and down bending stresses in the
shallow part of the subducted slab beneath the coast, are not incorporated in the
modeling here. The olivine-spinel phase change inside the slab (~ 400 km depth),
might give rise to important tensional and compressional stresses up to 1 GPa
(Goto et al., 1985). The maximum depth extent of the present models (Fig. 2) is
200 km, therefore the effect of olivine-spinel phase change on the stress
distribution of the subducting slab is not included here.
An important characteristic of the central Mexican subduction zone is its
wide shallow subhorizontal plate interface (~ 150 km) in the central part beneath
the Guerrero state and the sharp bending beneath the coast (Kostoglodov et al.,
1996). Another interesting distinctive feature for all the intraslab earthquakes in
central Mexico is the exclusively normal fault mechanism, they usually occurring for
distances starting ~ 85 km from the trench (Fig. 1).
Since the Cocos plate dips into the asthenosphere at an angle of ~ 20º due
to basalt to eclogite phase transformation (Manea et al., 2004a), pressure forces
(viscosity dependent) due to the induced flow in the mantle are partially balanced
by the gravitational body forces (density contrast dependent) (Turcotte and
Schubert, 2002).
- 168 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
This study aims to study the existence of exclusively normal fault intraslab
earthquakes in Mexico using the superposition of thermal stress and bending
stresses for the flat part of the subducted Cocos slab beneath Central Mexico.
- 169 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
MODELING PROCEDURE
Thermal Stresses
Using the temperature distribution of Manea et al., 2004b (Fig. 2), we
computed stresses in the subducting slab induced by thermal expansion. A system
of 2D equations for thin plates are solved for the Guerrero cross section (Fig. 1)
using the finite element solver PDE2D (http://pde2d.com/). The relevant equations
in explicit form are the two coupled linear steady state PDEs:
 ∂S11 ∂S12
 ∂x + ∂y = 0,


 ∂S12 + ∂S22 = 0, where:
 ∂x
∂y

(e11 + ν ⋅ e22 )
S11 = E ⋅ (1-ν 2 ) ,


(ν ⋅ e11 + e 22 )
S22 = E ⋅
and:
(1-ν 2 )


0.5 ⋅ E ⋅ e 12
,
S12 =
(1+ν )


∂u
− α ⋅ T ( x, y ),
e11 =
x
∂


∂v
e22 = − α ⋅ T ( x, y ), (1)
∂y


∂u ∂v
+ ,
e12 =
∂y ∂x

where:
(u,v)
-displacement vector,
T(x,y)
-2D temperature distribution,
E = 1.7*1012 dyne/cm2
-elastic modulus,
ν= 0.27
-Poisson ratio,
α= 10-5
-coefficient of thermal expansion.
The finite element grid extends from 20 km seaward of the trench up to 600
km landward. The lower limit of the model is considered the base of the oceanic
lithosphere and corresponds to the 1,150°C isotherm (Fig. 2). The upper limit of the
model is represented by the slab surface. The subducting plate presents two sharp
bending points beneath the coast (Kostoglodov et al., 1996), then follows a
subhorisontal segment and finally the slab submerges into the asthenosphere at a
- 170 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
dip angle of ~ 20º (Fig. 2). The elastic parameters are considered constant in the
numerical models. We use free boundary conditions (BC) at the tip of the
subducting slab and also for the circumference of the model. A fixed BC at the tip
of the slab increases the thermal stresses by ~ 50%, and following the conclusions
of Goto et al., 1983, free displacement at the tip gives a better explanation for the
seismicity. The grid consists of 5,000 triangle shaped elements (Fig. 3). The xcomponent (along the subducting slab) of the deviatoric stress is given by:
σ dx =
(2 ⋅ σ 'x -σ 'y )
3
(2)
,
where σ’x and σ’y are along the plate and normal to the plate thermal stresses
respectively. Because the thickness of the plate is much smaller that its length, the
σ’z is considered negligibly.
Bending Stresses
Since the Cocos plate dips into the asthenosphere at an angle of ~ 20º,
suction forces due to the induced flow in the mantle are partially balanced by the
gravitational body forces.
The pressure forces acting on the descending Cocos plate is calculated
using the 2D viscous corner flow model of Turcotte and Schubert, 2002. For an
oceanic plate that dips at 20° with a velocity of 5.5 cm/year, the following formula
for the pressure at the top of the plate is obtained:
P=
−8 ⋅ µ
⋅10−8
r
(3)
where:
- 171 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
P
-flow pressure on the top of the descending slab (Pa)
µ
-mantle wedge viscosity (Pa s)
r
-distance along the dipping slab surface with the origin in the hinge point
(m)
The negative value of P gives the effect of suction force, which tends to lift
up the slab. The lifting torque (MP) acting on the slab is:
MP = ∫
rmax
0
P ⋅ r ⋅ dr =∫
rmax
0
−8 ⋅ µ
⋅10−8 ⋅ r ⋅ dr = −8 ⋅ µ ⋅ rmax ⋅10−8 ( N ⋅ m)
r
(4)
where:
rmax
- the maximum plate length that contributes in transmitting the torque to
the hinge point.
The gravitational body force is calculated by:
Q = g ⋅ h ⋅ ∆ρ ⋅ cos(20o ) ( N / m)
(5)
where:
Q
-gravitational body force (N/m)
g
-gravitational acceleration (10 m/s2)
h
-elastic plate thickness (m)
∆ρ
-average density contrast over the elastic section of the slab:
∆ρ = ∆ρ ec log ite ⋅
hec log ite
helastic _ slab
+ ∆ρ peridotite ⋅
hperidotite
helastic _ slab
where:
- 172 -
(6)
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
∆ρeclogite = ρeclogite - ρperidotite = 3,500 – 3,300 = 200 kg/m3
∆ρperidotite = ρperidotite - ρperidotite (1 - α ∆T) = 3,300 – 3,300 (1 – 10-5 200°C)= 6.6 kg/m3
heclogite
-eclogite layer thickness (m)
hperidotite
-peridotite layer thickness (m)
helastic_slab = heclogite + hperidotite (m)
The pulling down torque (MQ) acting on the slab is:
MQ = ∫
rmax
0
r
Q ⋅ r ⋅ dr =Q ⋅ max
2
2
( N ⋅ m)
(7)
The net torque ( ∆M ) at the hinge point is:
∆M = M Q − M P
∆σ =
(8)
∆M
I
(9)
where:
∆σ
I=
b
-the bending stress due to ∆M
b ⋅ h3
12
-the moment of inertia (m4)
-the plate unity width (1m)
Choosing a positive (clockwise) net torque at the hinge point, the thermal
compressional stress at the top of the subhorisontal element of the slab (max. 0.65
kbars, see Fig. 4-B) might vanish.
- 173 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Stress distribution in the subducting slab.
Thermal stress in subduction zones arises due to the non-uniform
temperature distribution (Fig. 2). The results of the numerical calculations are
offered in Fig. 4, where along plate (Fig. 4-A) and deviatory along plate stresses
are presented (Fig. 4-B). The first part of the subducting plate (distances < 80 km
from the trench) is characterized by compressional deviatory stresses of ~ 0.3
kbars in its central part, while tensional deviatoric stresses occur in the upper and
lower part of the slab with a magnitude less than 0.3 kbars. For distances between
80 km and 115 km from the trench, the subducting plate presents a sharp bending,
this being a transition region from a compressional slab core to a tensional one.
Then there follows a subhorizontal sector where tensile thermal stresses are
representative for the core of the slab (max ~ 0.4 kbars) and compressional
stresses arise for the upper and lower parts of the slab (max. ~ 0.65 kbars close to
the hinge point). Afterward, the slab bends into the asthenosphere, and the same
pattern of the stresses is maintained, but with a greater magnitude of ~ 0.75 kbars
due to the elevated temperature contrast (~ 400°C) in this region.
The areas of tensional and compressional stresses are located parallel with
the dip of the plate. The maximum tensional stress (~ 0.75 kbars) occurs at a depth
of ~ 100 km. The compressional stress in the upper part of the plate is greater than
that for the lower part in about ~ 30%, these results being consistent with the
conclusion of Hamaguchi et al., 1983. Introducing for the lower part of the slab a
ductile behavior (Turcotte and Schubert, 2002), the bottom plate stresses reach a
maximum of 0.25 kbars, and then decrease exponentially to zero (see Fig. 5).
The high compressional thermal stress in the upper part of the slab is not
consistent with the exclusively normal fault intraslab earthquakes in central Mexico.
In a recent paper of Manea et al., 2004a, the hinge point (270 km from the trench)
is consistent with the transition of the basaltic oceanic crust to the eclogitic facies,
the significant increase in density bending the slab into the asthenosphere. The
cold slab dipping into the asthenosphere being denser than the surrounding
- 174 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
peridotite, contributes also to the gravitational torque through the elastic part of the
sinking slab. With a total density contrast of ∆ρ = 60 kg/m3, the elastic section of
the plate involved in gravitational bending is delimited by the ductile behavior,
therefore having a thickness of maximum 25 km (see Fig. 6 - inset). The net
downward torque at the hinge point due to the balance between suction and
gravitational forces (Fig. 6) might produce tensional stresses in the upper flat part
of the slab, and its magnitude might compensate the thermal compression (max. ~
0.65 kbars). In order to obtain tensile stresses at the top of the flat slab,
gravitational torque must be greater than the suction torque. Along the
subhorizontal part of the slab the suction force is constant and upward directed,
therefore, it does not contribute to the slab bending in this flat sector. The suction
torque depends on the mantle wedge viscosity and distance along dipping line,
while the gravity torque is function of moment of inertia (slab elastic thickness) and
distance along dipping line. Viscosities of 1018 Pa s up to 1019 Pa s are required for
the mantle wedge for slab lengths up to 500 km in order to fit a tensional stress of
0.65 kbars at the top of the flat part of the slab (see Fig. 7).
Flat slab seismicity and stress pattern
The intraslab seismicity in Guerrero is characterized by shallow and
intermediate depth intraslab normal fault earthquakes (Fig. 1 and Table 1). Most of
the events are confined in the subhorizontal part of the slab and a good correlation
with the deviatoric tensile thermal stress is observed (Fig. 5). Two clusters of
events, representing the background seismic activity with low magnitude, are
presented also (small circles in Fig. 2), and appear to be related with the sharp
bending-unbending of the plate in this region at ~ 80 km and ~ 110 km from the
trench. The maximum depth extent of the intraslab seismicity is ~ 80 km in Mexico,
being in agreement with the position of the 700ºC isotherm, which might be
considered as cutoff temperatures due to the transition from brittle to ductile
behavior.
- 175 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
The subhorizontal segment of the slab is the place where the majority of the
intraplate normal fault earthquakes occur. Thermal tensile (slab core) and
compressional (top, bottom) stresses arise in this region, but only normal events
are observed. Although thrust earthquakes have not been recorded, important
deviatoric compressional thermal stresses with magnitude of ~ 0.65 kbars occur in
the upper part of the slab. Introducing at hinge point a net downward torque (from
the balance of suction and gravity body forces) that gives 0.65 kbars tensile stress
at the top of the slab (Fig. 8-A), the top flat slab compression vanishes (Fig. 8-C).
In this case, a maximum tensile stress of ~ 0.55 kbars is obtained for the core of
flat slab. Between 700ºC - 800ºC isotherms the stress field in the slab twists from
tension to compression (Fig. 8-C), and the lack of the compressional seismicity is
well correlated with this cut off temperature range.
- 176 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
DISCUSSION AND CONCLUSIONS
An important individuality of the shallow and intermediate (depths of 35 - 80
km) intraslab Mexican earthquakes is their exclusively normal fault mechanism.
They usually occur at a distance greater than ~ 85 km from the trench. The
subhorizontal segment of the slab in Guerrero with a length of more than 150 km,
comprises the greater part of the normal fault intraslab events. Although the state
of stress in the subducting oceanic lithosphere is the superimposition of stress
distributions with various origins, including slab pull, ridge push, tidal stresses, etc.,
the present study focuses on the thermal and bending stresses. Based on the new
developed thermal models for Guerrero subduction zone (Manea et al., 2004b), the
thermal stress due to non-uniform temperature distribution in the subducting slab is
calculated using a finite element technique.
The first shallow part of the subducting slab is characterized by low
deviatoric compressional thermal stresses in its central part (~ 0.3 kbars) then the
stress pattern is reversed: a tensional behavior for the nucleus of the slab and a
compressional one for the upper and lower parts (Fig. 4-B). The core of the
subhorizontal part of the plate has a maximum value of thermal tension stress of ~
0.4 kbars, while the bottom of the slab reveals a thermal compression of ~ 0.75
kbars. Although intraslab reverse earthquakes have not been recorded, important
deviatoric compressional thermal stress with magnitude of ~ 0.65 kbars occurs in
the upper part of the flat slab. Introducing at the hinge point a net downward torque
that gives a tensile stress of 0.65 kbars at the top of the subhorizontal slab (Fig. 8A), a new stress field is obtained without compression in the upper flat segment
(Fig. 8-C). The lack of the compressional seimicity in the bottom part of the flat slab
is consistent with the position of the 700ºC - 800ºC isotherms, where the stress
filed toggle from tension to compression (see Fig. 8-C-inset). Also, the deficiency of
the intraslab normal fault earthquakes in Guerrero for distances less than ~85 km
from the trench is in good agreement with the onset of the tensional thermal stress
pattern in the core of the subducting slab.
- 177 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
REFERENCES
Conder, J.A., Weins, D.A., and Morris, J., 2002. On the decompression melting
structure at volcanic arcs and back-arc spreading centres. Geophysical Research
Letters, vol. 29, 17-1-17-4.
Dean, B.W., and Drake, C.L., 1978. Focal mechanism solutions and tectonics of
the Middle America arc. Journal of Geology, vol. 86, 111-128.
Engdahl, E. R. and Villaseñor, A., 2002. Global seismicity: 1900-1999, in Lee, W.
H. K. et al. (editors), International handbook of earthquake engineering and
seismology, International Geophysics series, vol. 81A, 665- 690,
Furukawa, F., 1993. Magmatic processes under arcs and formation of the volcanic
front. Journal of Geophysical Research, vol. 98, 8309-8319.
González-Ruiz, J.R., 1986. Earthquake source mechanics and tectonophysics of
the middle America subduction zone in Mexico, Ph.D. Thesis, Univ. of Calif., Santa
Cruz.
Gonzalez-Ruiz, J.R. and McNally, K.C., 1988. Stress accumulation and release
since 1882 in Ometepec, Guerrero, Mexico: implications for failure mechanisms
and risk assessments of a seismic gap, Journal of Geophysical Research, vol. 93,
6297-6317.
Gorbatov A., Kostoglodov, V., Suárez, G. and Gordeev, E., 1997. Seismicity
and structure of the Kamchatka subduction zone, Journal of Geophysical
Research, vol. 102, 17,883-17,898.
Goto, K., Hamaguchi, H., and Suzuki, Z., 1983. Distribution of stress in
descending plate in special reference to intermediate and deep focus earthquakes,
I. Characteristics of thermal stress distribution (the science reports of the Tohoku
University, Ser. 5), Tohoku Geophysics Journal, 29, 81-105.
Goto, K., Hamaguchi, H., and Suzuchi, Z., 1985. Earthquake generating stresses
in the descending slab. In; Kobayashi, K. and Sacks, I.S. (eds.), Structures and
Processes in the Subduction Zones. Tectonophysics, 112, 111-128.
- 178 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Hamaguchi, H., Goto, K., and Suzuki, Z., 1983. Double-planed structure of
intermediate-depth seismic zone and thermal stress in the descending plate,
Journal of Physics of the Earth, 331, 329-347.
Iglesias, A., Singh, S.K., Pacheco, J.F., and Ordaz, M., 2002. A Source and
Wave Propagation Study of the Copalillo, Mexico Earthquake of July 21, 2000
(Mw=5.9): Implications for Seismic Hazard in Mexico City from Inslab Earthquakes.
Bulletin of Seismological Society of America, v.92, no. 3, pp. 1060-1071.
Jiménez, Z. and Ponce, L., 1978. Focal mechanism of six large earthquakes in
northern Oaxaca, Mexico, of the period 1928-1973. Geofísica Internacional, vol.
17, 379-386.
Kelemen, P.B., Rilling, J.L., Parmentier, E.M., Mehl, L., and Hacker, B.R., 2003.
Thermal Structure due to Solid-State Flow in the Mantle Wedge Beneath Arcs, in
Inside the Subduction Factory, Geophysical Monograph 138, edited by J. Eiler,
AGU, Washington DC, pp. 293-311.
Kostoglodov, V., Bandy, W., Domínguez, J., and Mena, M., 1996. Gravity and
seismicity over the Guerrero seismic gap, Mexico.Geophysical Research Letters,
23, 3385-3388.
Lay. T., Astiz, L., Kanamori, H., and Christensen, D.H., 1989. Temporal
variation of large intraplate earthquakes in coupled subduction zones, Physics of
the Earth and Planetary Interiors, 54, 258-312.
Lefevre, L.V. and McNally, K.C., 1985. Stress distribution and subduction of
aseismic ridges in the Middle America subduction zone. Journal of Geophysical
Research, 90: 4495.
Manea, V.C., Manea, M., Kostoglodov, Currie, C.A., and Sewell, G., 2004a.
Thermal Structure, Coupling and Metamorphism in the Mexican Subduction Zone
beneath Guerrero. Geophysical Journal International, 158, 775–784 doi:
10.1111/j.1365 - 246X.2004.02325.x.
Manea, V.C., Manea, M., Kostoglodov, V., Sewell, G., 2004b. (in press)
Thermo-mechanical model of the mantle wedge in Central Mexican subduction
- 179 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
zone and a blob tracing approach for the magma transport. Physics of the Earth
and Planetary Interiors.
Pardo, M., and Suárez, G., 1995. Shape of the subducted Rivera and Cocos
plates in southern Mexico: seismic and tectonic implications. Journal of
Geophysical Research, 100, 12357-12373.
Singh, S.K., Suárez, G., and Domínguez, T., 1985. The great Oaxaca earthquake
of 15 January 1931: lithosphere normal faulting in the subducted Cocos plate.
Nature, 317, 56-58.
Singh, S.K., and Pardo, M., 1993. Geometry of the Benioff zone and the state of
stress in the overriding plate in central México. Geophysical Research Letters, 20,
1483-1486.
Turcotte, D.L., and Schubert, G., 2002. Geodynamics, 2nd edition. Cambridge
University Press, New York.
van Hunen, J., van den Berg, A., and Vlaar, N., 2002. On the role of subducting
oceanic plateaus in the development of shallow flat subduction. Tectonophysics,
352, 317-333.
van Keken, P.E., Kiefer, B., and Peacock, S.M., 2002. High resolution models of
subduction zones: Implications for mineral dehydration reactions and the transport
of water into deep mantle. G-cubed, 3, 10, 20.
- 180 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Table 1.
Moderate and large normal fault intraslab events in the subducted Cocos
plate for the region of interest (Fig. 1) (15 - 200N and 96 - 1040W). If magnitude
4
5
6
7
8
9
10
1959.05.24
1964.07.06
1966.02.27
1966.09.25
1967.04.13
1967.04.14
1968.08.14
17.72
18.31
18.82
18.12
18.12
17.2
18.28
97.15
100.5
102.56
100.96
100.4
100.43
103.05
11 1971.03.03
17.61
99.37
12 1971.07.16
16.24
96.34
13 1971.10.27
14 1973.07.03
15 1973.08.28
18.07
18.85
18
100.52
100.03
96.55
16
17
18
19
20
21
22
23
1974.01.26
1974.07.18
1976.09.05
1976.09.19
1978.07.05
1978.09.29
1983.12.08
1984.06.04
18.97
17.1
18.38
17.98
18.74
18.3
18.36
17.62
103.84
98.4
101.36
100.65
100.17
102.46
102.73
98.03
24 1985.07.04
17.55
97.08
- 181 -
Reference
96.87
97.65
Rake
16.34
18.32
Dip
2 1931.01.15
3 1945.10.11
84 6.5 MGR 343 70 -117 Jiménez and Ponce (1978)
40
7.8
270 56 -90 Singh et al. (1985)
95 6.5 MGR 343 65 -61 Jiménez and Ponce (1978)
80
6.8
315 61 -102 Jiménez and Ponce (1978)
55
7.3
292 38 -63 González-Ruiz et al. (1986)
89
5.8
165 35 -107 Lefebevre & McNally (1982)
59
5.8
290 50 -84 Pardo and Suárez (1995)
65
5.5
314 56 -92 Pardo and Suárez (1995)
35
5
160 48 -106 Singh & Pardo (1993)
44
5.4
319 84 -40 Lay et al. (1989), Dean &
Drake (1978), Engdahl and
Villaseñor(2002)
73
5.1
298 56 -90 Dean & Drake (1978), Lay et
al. (1989), Engdahl and
Villaseñor(2002)
49
5.3
290 45 -60 Cruz & Suárez (Personal
comunication)
59
5
326 60 -90 Mota (1973)
95
5.9
281 80 -98 Pardo and Suárez (1995)
82 7.3 mb 326 50 -75 González-Ruiz et al. (1986),
González-Ruiz & McNally
(1988)
51 5.2 mb 289 59 -30 Pardo and Suárez (1995)
48 5.6 mb 243 63 174 Lefevre & McNally (1985)
63 5.4 mb 297 64 -102 Pardo and Suárez (1995)
62
5.5
295 54 -66 Pardo and Suárez (1995)
59
5.5
311 47 -92 Pardo and Suárez (1995)
63
5.5
281 33 -80 Pardo and Suárez (1995)
54
5.6
262 40 -90 Pardo and Suárez (1995)
62
5.3
289 45 -72 CMT Centroid Moment
Tensor (Harvard catalog)
68
5.1
318 41 -98 CMT (Harvard catalog)
Azimuth
Depth (km)
97.99
Magnitude
Longitude
(°W)
18.26
Date
1 1928.02.10
No. crt.
Latitude
(°N)
type is not given then M = Mw.
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
25 1987.07.15
17.334
97.31
66
5.8
26
27
28
29
30
31
32
33
34
35
36
37
38
39
18.56
18.07
17.62
18.22
18.19
17.26
18.55
18.42
16.42
18.76
18.44
18.09
19.16
18.113
101.4
101.13
97.94
96.82
100.2
96.1
100.8
101.84
96.06
101.73
97.38
101.62
103.23
98.974
74
47
54
75
70
40
78
59
57
56
61
48
63
50
5.1
5.3
5.2
5.8
6.2
5.5
5.3
5.5
5.2
6.5
6.9
6.2
5.2
5.8
1987.07.26
1989.08.12
1993.08.05
1994.02.23
1994.05.23
1994.08.28
1995.12.20
1996.01.25
1996.04.01
1997.05.22
1999.06.15
1999.06.21
1999.08.15
2000.21.06
338 42 -65 Cruz & Suárez (Personal
comunication)
280 68 -93 CMT (Harvard catalog)
270 71 -90 Pardo and Suárez (1995)
294 39 -135 CMT (Harvard catalog)
278 36 -83 CMT (Harvard catalog)
273 39 -76 CMT (Harvard catalog)
302 72 -89 CMT (Harvard catalog)
274 61 -77 CMT (Harvard catalog)
276 34 -72 CMT (Harvard catalog)
306 64 -111 CMT (Harvard catalog)
269 63 -96 CMT (Harvard catalog)
309 40 -83 CMT (Harvard catalog)
296 32 -88 CMT (Harvard catalog)
295 74 -104 CMT (Harvard catalog)
305 32 -80 Iglesias et al. (2002)
- 182 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
FIGURE CAPTIONS
Figure 1.
The position of the modeled cross-section in Guerrero. Focal mechanisms
represent the intraslab normal fault events with magnitude greater than 5.0 (Mw).
The thick grey line represents the 2D model cross-section. MAT - Middle American
Trench.
Figure 2.
Thermal model for Guerrero with strongly temperature dependence of the
viscosity from Manea et al., 2004b. The focal mechanisms represent the intraplate
seismicity with magnitude greater than 5.0 (Mw) from Fig. 1.
Figure 3.
A mesh with more than 5,000 triangles is used in thermal stresses
computation.
Figure 4.
(A). Thermal stresses along the subducting slab.
(B). Deviatoric thermal stresses along the subducting slab.
Figure 5.
Deviatoric thermal stresses with ductile behavior included (Turcotte and
Schubert, 2002) at the bottom of the slab. The inset represents a cross-section
through the slab at 200 km from the trench.
Figure 6.
Induced flow pressure and gravitational body force acting on the descending
slab.
- 183 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 7.
Mantle wedge viscosities inferred from the net torque (∆σ = 0.65 kbars) at
the hinge point.
Figure 8.
Superimposition (C) of the bending (A) and thermal stresses (B) for the flat
part of the subducting slab. The insets represent corss-sections through the slab at
200 km from the trench.
- 184 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 1
- 185 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 2
- 186 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 3
- 187 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 4
- 188 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 5
- 189 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 6
- 190 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 7
- 191 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Figure 8
- 192 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
VI. DISCUSIÓN Y CONCLUSIONES
Las estructuras térmicas 2D de las zonas de subducción de la parte central
de México y del sur de Kamchatka han sido estudiadas resolviendo un sistema de
ecuaciones de Stokes y del flujo de calor con la ayuda de los elementos finitos.
Los modelos numéricos de la distribución de temperatura en el antearco de
la parte central de la zona de subducción de Guerrero fueron constreñidos por: los
datos de flujo de calor reportados para la superficie; la forma de la interfaz de la
placa subducida deducida de los modelos de gravedad, datos de sismicidad y
estimaciones recientes de la extensión del acoplamiento a lo largo de la zona de
contacto entre las placas de Cocos y Norteamérica. La zona sismogénica
modelada, delimitada por las isotermas de 100 ºC – 150 ºC y 250 ºC, coincide con
el ancho de la zona de ruptura de los terremotos (subduction thrust earthquakes)
en Guerrero. Una pequeña cantidad de calor generada por fricción es necesaria
para ajustar la extensión máxima de la zona parcialmente acoplada (la isoterma de
450 ºC), de acuerdo con el ancho de la zona acoplada calculado a través de los
modelos de deformación. Para ajustar la posición de la isoterma de 450 ºC a una
distancia de ~ 200 km de la trinchera, que corresponde a una extensión máxima
del acoplamiento entre las placas, es necesario tener una interfaz subhorizontal de
la placa subducida entre 115 y 270 km desde la trinchera.
El cambio en las secuencias metamórficas en la corteza oceánica
subducida se relaciona con la variación del acoplamiento a lo largo del contacto
entre las placas, estimado por las observaciones de las deformaciones en
superficie durante el periodo intersísmico. Para el rango de temperaturas de 250ºC
- 450 ºC y presiones de 0.6 - 1.3 GPa las facies metamórficas están representadas
por jadeita-lawsonita-esquitos azules y epidot-esquitos azules. Los esquitos azules
y las facies asociadas en estos rangos de presión y temperatura probablemente se
comportan de manera
dúctil. Este comportamiento es responsable del
acoplamiento parcial de largo plazo y los eventos asísmicos esporádicos. Para
temperaturas mayores a 200 ºC - 300 ºC las rocas preexistentes sufren cambios
- 193 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
marcados en su textura y mineralogía. La presión y el calor son los agentes
principales del metamorfismo. El efecto del calor sobre el basalto de la placa
oceánica aumenta la ductilidad y el cambio en la composición mineralógica. Las
rocas metamórficas foliadas, como los esquites azules, presentan una textura
aplanada que favorece el comportamiento dúctil a lo largo de la interfaz de
contacto entre las placas. La deformación dúctil que resulta de los esfuerzos de
cizalla, que caracterizan a la zona de subducción, es uno de los responsables del
desarrollo de la anisotropía en las rocas metamórficas, como en los esquitos
azules.
Nuestros modelos muestran una buena correlación entre la posición del
punto de bisagra (270 km desde la trinchera) y la ocurrencia de las facies de
eclogita en la corteza oceánica de la placa subducida. En estos modelos, la
temperatura de 250ºC coincide con la extensión máxima de la zona sismogénica
(~ 82 km desde la trinchera) y con la existencia de las facies de esquitos azules y
de las facies asociadas.
De acuerdo con el diagrama de fases para el MORB, la deshidratación
intensiva de la corteza oceánica en subducción debe ocurrir para T = 250 ºC 450ºC y P = 0.6 - 1.3 GPa, liberándose más de 2 wt% H2O durante el cambio de
fase. La ocurrencia de los eventos lentos puede estar relacionada con esta
deshidratación.
Las estructuras térmicas debajo de las fajas volcánicas del México central y
del sur de Kamchatka dependen de los parámetros reológicos, la viscosidad de
referencia, η0, y la energía de activación, Ea.
Para México, el diagrama de fases para minerales máficos muestra que el
fundido de la corteza oceánica basaltica puede ocurrir en las profundidades de ~
58 km para η0 = (1017 - 1020 Pa s) y de ~ 80 km para η0 = 1021 Pa s. Los
sedimentos saturados en agua empiezan a fundirse a profundidades más
someras, entre 50 y 55 km. El perfil vertical de la temperatura debajo del volcán
Popocatépetl muestra temperaturas bastante grandes para fundir la peridotita
hidratada.
- 194 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
En el rango de valores de viscosidad de referencia (1017 - 1021 Pa s), los
sedimentos saturados pueden fundir a profundidades entre 45 y 50 km. La
temperatura debajo del volcán Popocatépetl es superior a la del solidus de la
peridotita hidratada para todo el rango de las energías de activación para olivino
usadas en este estudio (150 - 350 kJ/mol).
La temperatura debajo del CVTM es superior la del solidus para peridotita
hidratada, pero es más baja para la peridotita deshidratada. Esto sugiere que la
hidratación de la cuña del manto por fluidos liberados de la placa subducida de
Cocos es una condición necesaria para producir fundido de la peridotita del manto.
Esto está de acuerdo con los resultados de nuestros modelos que muestran que la
deshidratación de las rocas metamorfizadas puede ocurrir a profundidades hasta
de ~ 80 km. Debido a la transformación de la zoicita y anfíbol en eclogita, ~ 0.6
wt% H2O puede ser liberado en la cuña del manto para profundidades entre 40 60 km.
Los xenolitos del manto han sido encontrados en México cerca de El
Peñón, sugiriendo un aporte importante de volátiles de la placa subducida de
Cocos. Las rocas de composición calcialcalinas son las más comunes en el
CVTM. Esta depleción ocurre probablemente debido a la cristalización de los
oxidos de Fe y Ti, que inicialmente puede ser favorecida por la presencia de
fluidos en el magma. La gran mayoría de las rocas calcolcalinas en el CVTM está
representada por rocas con gran contenido en K2O y Na2O, que originalmente
pueden representar el fundido parcial en la cuña del manto.
El magma con una composición félsica ha sido encontrado en el campo de
volcanes mono-génicos de Chichinautzin, sugiriendo el fundido parcial de la
corteza basáltica. Esto esta de acuerdo con los modelos de temperatura que
muestran temperaturas > 1100 ºC en la base de la corteza continental.
El modelo de transporte del magma usando el movimiento de una burbuja
flotante en la cuña del manto nos puede proporcionar una herramienta sencilla
para estudiar el origen del volcanismo en el CVTM. Las burbujas flotantes pueden
tener su origen en la interfaz entre la placa subducida y la cuña del manto como
- 195 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
una consecuencia de la inestabilidad térmica y/o química. Una cierta distancia a lo
largo de la interfaz es necesaria para que el magma se acumule formando una
burbuja flotante. Sin duda, en este modelo la zona de salida de las burbujas ha
sido elegido justamente debajo (a una profundidad de 70 km) de la estructura
volcánica principal (el volcán Popocatépetl). Las burbujas con un diámetro de 10
km pueden alcanzar la base de la corteza continental en ~ 6 Ma, para una
viscosidad de ~ 3·1017 Pa s. El tiempo necesario para una burbuja de 1 km en
diámetro para alcanzar la corteza continental varia entre 1000 años y 14 Ma, para
viscosidades entre 1014 Pa s y 5·1017 Pa s. Por lo general, el tiempo de subida
para una burbuja flotante diminuye mientras el diámetro aumenta y la viscosidad
disminuye.
El modelo dinámico de la trayectoria de la burbuja flotante en la cuña del
manto muestra que las trayectorias más “rápidas” terminan en el mismo punto
(debajo del volcán Popocatépetl, ~ 350 km desde la trinchera) a la base de la
corteza continental. Este resultado puede ser interpretado como una posible
condición de desarrollo para los estratovolcanes.
Estudios recientes muestran por lo menos dos pulsos magmáticos en un
lapso de tiempo de ~ 1Ma en el volcán Nevado de Toluca. El volumen máximo de
material extrudido para cada uno de estos pulsos es de ~ 3.5 km3 (equivalente a
una burbuja de ~ 2 km en diámetro). Los dos pulsos presentan una firma
magmática de sedimentos fundidos (la anomalía de Ce). De acuerdo con los
modelos del capitulo III de este estudio, una burbuja de 2 km diámetro alcanza la
base de la corteza continental en ~ 1 Ma para una viscosidad de ~ 2·1015 Pa s.
Esta viscosidad muy baja es esencial para que las burbujas alcancen la base de la
corteza continental.
El volumen promedio de los conos monogéneticos del CVTM es menor al 1
km3, valor que corresponde a un diámetro de la burbuja de ~ 1.3 km. Si el origen
de los conos monogéneticos es descrito por el modelo de la trayectoria de la
burbuja flotante, la viscosidad debería ser η > 5·1015 Pa s para producir las
trayectorias “lentas”.
- 196 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Una característica principal de los terremotos intraplaca en la zona de
subducción mexicana (para terremotos someros e intermedios, con profundidades
entre 35 y 80 km) es su mecanismo focal normal. Estos terremotos ocurren
generalmente a distancias de ~ 85 km desde la trinchera. En Guerrero hay un
segmento de la placa subducida de forma subhorizontal, con una longitud de ~
150 km, que contiene la mayoría de los eventos normales intraplaca. Aunque el
estado de esfuerzos en la litósfera oceánica subducida es un conjunto de
esfuerzos de varios orígenes, como el slab-pull, ridge-push, esfuerzos de mareas,
etc., en este estudio se trata sólo de los esfuerzos térmicos y torsión. Usando los
nuevos modelos térmicos desarrollados en el capitulo III de este estudio para la
zona de Guerrero, los esfuerzos térmicos producidos por una distribución nonuniforme de la temperatura en la placa subducida de Cocos, son calculados
usando la técnica de los elementos finitos.
Los resultados muestran que la primera parte somera de la placa subducida
está caracterizada por esfuerzos termoelásticos desviatorios compresionales de
magnitud baja (~ 0.3 kbars) en su parte central. Luego, el patrón de los esfuerzos
térmicos se cambia: un comportamiento extensional para el centro de la placa y
uno compresional para las partes de superior e inferior de la placa. El núcleo de la
parte subhorizontal de la placa subducida tiene un valor máximo de esfuerzo
termoelástico de tensión de ~ 0.4 kbars, mientras para la parte inferior de la placa,
muestra unos esfuerzos termoelásticos de extensión de ~ 0.75 kbars. Incluyendo
una dependencia entre los esfuerzos y la temperatura alta para la parte inferior de
la placa, los esfuerzos termoelásticos de compresión de ~ 0.75 kbars desaparecen
debido al comportamiento dúctil de la placa. Aunque los terremotos intraplaca con
mecanismo inverso no han sido registrados, hay unos esfuerzos termoelásticos
desviatorios compresionales con magnitud de ~ 0.65 kbars en la parte superior de
la parte horizontal de la placa subducida.
Incluyendo al punto de bisagra un
momento de torsión en el sentido de las agujas del reloj que puede producir
esfuerzos de tensión de 0.65 kbars en la parte superior de la placa subducida subhorizontal, se obtiene un nuevo campo de esfuerzos, sin compresión en la parte
- 197 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
superior de este segmento. La falta de la sismicidad compresional en la parte
inferior de la placa subducida así como la profundidad máxima de estos
terremotos, de ~ 80 km, está de acuerdo con la posición de las isotermas de
700ºC - 800ºC. Para temperaturas mayores a este intervalo, la placa tiene un
comportamiento dúctil y como consecuencia no pueden ocurrir más terremotos
intraplaca.
El mismo esquema numérico utilizado para calcular los modelos termomecánicos para Guerrero, se usó para otra zona de subducción muy activa y
antigua, la zona de subducción del sur de Kamchatka.
Los resultados muestran una temperatura > 1,300 ºC y dos posibles fuentes
para el fundido: los sedimentos y la peridotita de la cuña del manto debajo del
frente volcánico. Una pequeña cantidad de calor producido por fricción (un
coeficiente efectivo de fricción de µ = 0.034 o un esfuerzo cortante τ ~ 14 MPa)
está añadida al contacto entre la Placa de Pacífico y la Península de Kamchatka.
Para cantidades mayores de calor producido por fricción, el fundido de la corteza
oceánica basáltica puede ocurrir, pero esto no esta de acuerdo con la falta de
magmas adakíticos en el Sur de Kamchatka. El modelo térmico muestra un patrón
de velocidad en la cuña del manto debajo del Moho
que puede inducir una
anisotropía del olivino. Esta anisotropía está de acuerdo con los estudios que
muestran que los ejes anisotrópicós rápidós son perpendiculares a la trinchera en
el Sur de Kamchatka.
Una deshidratación mayor de la corteza oceánica basáltica (> 5 wt% H2O
liberado) se produce debajo del arco volcánico hasta una profundidad de ~ 100
km. Como consecuencia, justamente arriba de la placa subducida (~ 5 km) puede
existir una capa de peridotita y sedimentos fundidos. Esto puede producir
instabilidades térmicas y/o químicas que se pueden acumular en forma de
burbujas muy cerca de la superficie de la placa subducida. Basado en esto, el
modelo dinámico de la trayectoria de las burbujas en el campo de velocidades de
la cuña del manto se aplica también para el sur de Kamchatka.
- 198 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
Para constreñir el tiempo para la subida del magma a través del manto se
usan los estudios sobre los desequilibrios de las series isotópicas de U,
10
Be y las
trazas de los elementos solubles como los isótopos de Th, Sr, y Pb. Los estudios
de los desequilibrios de las series isotópicas de U proponen una subida muy
rápida del magma (~ 1000 años
~ 60 m/a) desde la superficie de la placa
subducida hacia la superficie a través de una red de canales. Otros estudios dan
una velocidad del magma < 350 ka (o ~ 17 cm/a), deducida por los datos de los
isótopos de Th, Sr, Pb. Otros estudios recientes para los desequilibrios de U-ThPa-Ra para Kamchatka discuten sobre la existencia de un modelo dinámico de
fundido que no requiere una velocidad muy alta de subida (~ 1 m/a) dentro de la
cuña del manto. Una conclusión más contradictoria proviene de los estudios del
10
Be que muestran un tiempo de residencia de los isótopos de Be en la cuña del
manto de ~ 6 Ma antes de salir a la superficie como actividad volcánica.
Los desequilibrios de los isótopos de U-Th-Pa-Ra, Th, Sr y Pb asimismo
como los estudios de
10
Be muestran estimaciones muy contradictorias para los
tiempos necesarios para el transporte de magma en el manto desde < 350 ka
hasta 6 Ma. Esta variabilidad tan alta en el transporte del magma sugiere tasas
muy variables en el transporte de varios elementos y/o diferencias reales en los
tiempos de transporte probablemente debido a volúmenes variables de fundido en
el manto.
El modelo de las burbujas flotantes, a pesar de su simplicidad, tiene la
ventaja de explicar un gran rango de tiempos de residencia del magma en la cuña
del manto.
En este estudio, se predice la historia térmica de una burbuja flotante de 10
km de diámetro, que se mueve a través de la cuña del manto en un campo
térmico, usando las ecuaciones de la conducción del calor. Los resultados
muestran que después de ~ 1 Ma, la burbuja “fría” (~ 800ºC), con un diámetro de
10 km, se mueve hacia la región más “caliente” de la cuña del manto, en donde
hay temperaturas de más de 1300ºC. Después de calentarse, la burbuja se mueve
hacia arriba hasta la base de la corteza continental, en donde llega con un núcleo
- 199 -
Manea, V.C. – Modelos termomecánicos para las zonas de subducción de Guerrero y Kamchatka – 2004
con temperaturas > 900ºC. De acuerdo con el solidus para peridotita hidratada la
mayor parte del volumen de la burbuja de 10 km (T > 900ºC a 1 GPa) podría ser
fundido al alcanzar el Moho.
Finalmente, en este estudio, el modelo térmico para el sur de Kamchatka se
usa para estimar la anomalía de velocidad en la cuña del manto. Las grandes
temperaturas en la cuña del manto debajo del arco volcánico (> 1300ºC) producen
una anomalía de velocidad negativa muy fuerte de hasta –7% (con respecto al
PREM). Por otra parte, la subducción de la placa oceánica fría produce anomalías
de velocidad positivas de hasta +4%. La buena correlación que hay entre las
perturbaciones en la velocidad debajo del arco volcánico (por lo menos en la
magnitud) entre la imagen tomográfica de las llegadas de las ondas P (-7%) y las
estimaciones de las anomalías de velocidad de este estudio (~-7%), sugiere que el
modelado del campo térmico para la cuña del manto en el sur de Kamchatka ha
sido satisfactorio.
- 200 -